{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 23583, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00012721027859051011, "ewc_loss": 0.0, "ewc_loss_diag": 0.0, "ewc_loss_parallel": 0.0, "grad_norm": 4.834580421447754, "learning_rate": 0.0, "loss": 0.7982, "mean_token_accuracy": 0.7762961387634277, "num_tokens": 38493.0, "step": 1 }, { "epoch": 0.00025442055718102023, "ewc_loss": 0.0, "ewc_loss_diag": 0.0, "ewc_loss_parallel": 0.0, "grad_norm": 4.58832311630249, "learning_rate": 4.2390843577787196e-10, "loss": 0.8329, "mean_token_accuracy": 0.765798807144165, "num_tokens": 80419.0, "step": 2 }, { "epoch": 0.0003816308357715303, "ewc_loss": 4.85722573273506e-15, "ewc_loss_diag": 6.132518538121134e-19, "ewc_loss_parallel": 4.797594613248357e-18, "grad_norm": 4.726110935211182, "learning_rate": 8.478168715557439e-10, "loss": 0.7225, "mean_token_accuracy": 0.7960126996040344, "num_tokens": 118717.0, "step": 3 }, { "epoch": 0.0005088411143620405, "ewc_loss": 2.0161650127192843e-13, "ewc_loss_diag": 5.0523821237824507e-17, "ewc_loss_parallel": 1.960237527853792e-16, "grad_norm": 5.3669753074646, "learning_rate": 1.271725307333616e-09, "loss": 0.8139, "mean_token_accuracy": 0.7712426781654358, "num_tokens": 150155.0, "step": 4 }, { "epoch": 0.0006360513929525506, "ewc_loss": 1.7053025658242404e-12, "ewc_loss_diag": 1.6930901125533637e-15, "ewc_loss_parallel": 1.5334955527634975e-15, "grad_norm": 4.3047637939453125, "learning_rate": 1.6956337431114878e-09, "loss": 0.7917, "mean_token_accuracy": 0.7747752666473389, "num_tokens": 193616.0, "step": 5 }, { "epoch": 0.0007632616715430606, "ewc_loss": 8.86757334228605e-12, "ewc_loss_diag": 5.384581669432009e-15, "ewc_loss_parallel": 8.271161533457416e-15, "grad_norm": 5.21329927444458, "learning_rate": 2.1195421788893596e-09, "loss": 0.7894, "mean_token_accuracy": 0.7782890200614929, "num_tokens": 227640.0, "step": 6 }, { "epoch": 0.0008904719501335708, "ewc_loss": 1.716671249596402e-11, "ewc_loss_diag": 1.1213252548714081e-14, "ewc_loss_parallel": 1.5987211554602254e-14, "grad_norm": 4.835421562194824, "learning_rate": 2.543450614667232e-09, "loss": 0.8162, "mean_token_accuracy": 0.7746796607971191, "num_tokens": 265114.0, "step": 7 }, { "epoch": 0.001017682228724081, "ewc_loss": 7.594280759803951e-11, "ewc_loss_diag": 3.397282455352979e-14, "ewc_loss_parallel": 7.283063041541027e-14, "grad_norm": 4.932100772857666, "learning_rate": 2.967359050445104e-09, "loss": 0.7584, "mean_token_accuracy": 0.7883440256118774, "num_tokens": 299865.0, "step": 8 }, { "epoch": 0.001144892507314591, "ewc_loss": 1.2096279533579946e-10, "ewc_loss_diag": 5.1958437552457326e-14, "ewc_loss_parallel": 1.1546319456101628e-13, "grad_norm": 4.565450191497803, "learning_rate": 3.3912674862229757e-09, "loss": 0.8138, "mean_token_accuracy": 0.7738631963729858, "num_tokens": 342063.0, "step": 9 }, { "epoch": 0.0012721027859051012, "ewc_loss": 2.2646418074145913e-10, "ewc_loss_diag": 1.2878587085651816e-13, "ewc_loss_parallel": 2.1405099914773018e-13, "grad_norm": 5.45689582824707, "learning_rate": 3.815175922000847e-09, "loss": 0.864, "mean_token_accuracy": 0.7641929388046265, "num_tokens": 374864.0, "step": 10 }, { "epoch": 0.0013993130644956112, "ewc_loss": 6.439222488552332e-10, "ewc_loss_diag": 2.220446049250313e-13, "ewc_loss_parallel": 6.217248937900877e-13, "grad_norm": 4.298281669616699, "learning_rate": 4.239084357778719e-09, "loss": 0.7765, "mean_token_accuracy": 0.7752736210823059, "num_tokens": 416605.0, "step": 11 }, { "epoch": 0.0015265233430861213, "ewc_loss": 8.731149137020111e-10, "ewc_loss_diag": 3.232969447708456e-13, "ewc_loss_parallel": 8.419931418757187e-13, "grad_norm": 5.442785263061523, "learning_rate": 4.662992793556591e-09, "loss": 0.8349, "mean_token_accuracy": 0.7692875862121582, "num_tokens": 448798.0, "step": 12 }, { "epoch": 0.0016537336216766315, "ewc_loss": 1.0986695997416973e-09, "ewc_loss_diag": 3.907985046680551e-13, "ewc_loss_parallel": 1.0658141036401503e-12, "grad_norm": 5.43853235244751, "learning_rate": 5.086901229334464e-09, "loss": 0.8739, "mean_token_accuracy": 0.7539591789245605, "num_tokens": 480084.0, "step": 13 }, { "epoch": 0.0017809439002671415, "ewc_loss": 3.448803909122944e-09, "ewc_loss_diag": 1.0729195309977513e-12, "ewc_loss_parallel": 3.353761712787673e-12, "grad_norm": 4.376841068267822, "learning_rate": 5.510809665112336e-09, "loss": 0.8057, "mean_token_accuracy": 0.7794235944747925, "num_tokens": 524543.0, "step": 14 }, { "epoch": 0.0019081541788576518, "ewc_loss": 5.093170329928398e-09, "ewc_loss_diag": 1.5916157281026244e-12, "ewc_loss_parallel": 4.945377440890297e-12, "grad_norm": 4.776602268218994, "learning_rate": 5.934718100890208e-09, "loss": 0.7518, "mean_token_accuracy": 0.7880719304084778, "num_tokens": 563314.0, "step": 15 }, { "epoch": 0.002035364457448162, "ewc_loss": 6.315531209111214e-09, "ewc_loss_diag": 1.7692514120426495e-12, "ewc_loss_parallel": 6.139089236967266e-12, "grad_norm": 5.083012580871582, "learning_rate": 6.3586265366680796e-09, "loss": 0.8425, "mean_token_accuracy": 0.762475848197937, "num_tokens": 598421.0, "step": 16 }, { "epoch": 0.002162574736038672, "ewc_loss": 7.741618901491165e-09, "ewc_loss_diag": 2.2168933355715126e-12, "ewc_loss_parallel": 7.503331289626658e-12, "grad_norm": 5.038357734680176, "learning_rate": 6.782534972445951e-09, "loss": 0.8321, "mean_token_accuracy": 0.7626471519470215, "num_tokens": 634690.0, "step": 17 }, { "epoch": 0.002289785014629182, "ewc_loss": 8.672941476106644e-09, "ewc_loss_diag": 2.6574298317427747e-12, "ewc_loss_parallel": 8.355982572538778e-12, "grad_norm": 5.123055458068848, "learning_rate": 7.206443408223823e-09, "loss": 0.8691, "mean_token_accuracy": 0.7600266933441162, "num_tokens": 674653.0, "step": 18 }, { "epoch": 0.0024169952932196924, "ewc_loss": 2.3981556296348572e-08, "ewc_loss_diag": 9.947598300641403e-12, "ewc_loss_parallel": 2.2964741219766438e-11, "grad_norm": 5.063526153564453, "learning_rate": 7.630351844001695e-09, "loss": 0.8005, "mean_token_accuracy": 0.7767610549926758, "num_tokens": 708238.0, "step": 19 }, { "epoch": 0.0025442055718102024, "ewc_loss": 3.5390257835388184e-08, "ewc_loss_diag": 1.2107648217352107e-11, "ewc_loss_parallel": 3.433342499192804e-11, "grad_norm": 4.602728843688965, "learning_rate": 8.054260279779567e-09, "loss": 0.7825, "mean_token_accuracy": 0.7787966728210449, "num_tokens": 749312.0, "step": 20 }, { "epoch": 0.0026714158504007124, "ewc_loss": 4.1676685214042664e-08, "ewc_loss_diag": 1.4097167877480388e-11, "ewc_loss_parallel": 4.0245140553452075e-11, "grad_norm": 5.412014484405518, "learning_rate": 8.478168715557438e-09, "loss": 0.8525, "mean_token_accuracy": 0.7664927840232849, "num_tokens": 783532.0, "step": 21 }, { "epoch": 0.0027986261289912225, "ewc_loss": 4.6566128730773926e-08, "ewc_loss_diag": 1.4779288903810084e-11, "ewc_loss_parallel": 4.524736141320318e-11, "grad_norm": 5.065722942352295, "learning_rate": 8.902077151335311e-09, "loss": 0.83, "mean_token_accuracy": 0.7735228538513184, "num_tokens": 817429.0, "step": 22 }, { "epoch": 0.0029258364075817325, "ewc_loss": 5.471520125865936e-08, "ewc_loss_diag": 1.77351466845721e-11, "ewc_loss_parallel": 5.275069270282984e-11, "grad_norm": 4.8199896812438965, "learning_rate": 9.325985587113182e-09, "loss": 0.7633, "mean_token_accuracy": 0.7887709736824036, "num_tokens": 853964.0, "step": 23 }, { "epoch": 0.0030530466861722425, "ewc_loss": 5.8906152844429016e-08, "ewc_loss_diag": 1.9326762412674725e-11, "ewc_loss_parallel": 5.707079253625125e-11, "grad_norm": 5.465689659118652, "learning_rate": 9.749894022891054e-09, "loss": 0.8282, "mean_token_accuracy": 0.7669211626052856, "num_tokens": 885070.0, "step": 24 }, { "epoch": 0.003180256964762753, "ewc_loss": 6.379559636116028e-08, "ewc_loss_diag": 2.1373125491663814e-11, "ewc_loss_parallel": 6.139089236967266e-11, "grad_norm": 4.8312907218933105, "learning_rate": 1.0173802458668929e-08, "loss": 0.8799, "mean_token_accuracy": 0.7511045336723328, "num_tokens": 926893.0, "step": 25 }, { "epoch": 0.003307467243353263, "ewc_loss": 1.434236764907837e-07, "ewc_loss_diag": 6.775735528208315e-11, "ewc_loss_parallel": 1.3733369996771216e-10, "grad_norm": 4.875331401824951, "learning_rate": 1.05977108944468e-08, "loss": 0.8159, "mean_token_accuracy": 0.7676070332527161, "num_tokens": 964773.0, "step": 26 }, { "epoch": 0.003434677521943773, "ewc_loss": 2.2351741790771484e-07, "ewc_loss_diag": 8.685674401931465e-11, "ewc_loss_parallel": 2.15550244320184e-10, "grad_norm": 4.741640090942383, "learning_rate": 1.1021619330224672e-08, "loss": 0.769, "mean_token_accuracy": 0.7853063344955444, "num_tokens": 1002725.0, "step": 27 }, { "epoch": 0.003561887800534283, "ewc_loss": 2.60770320892334e-07, "ewc_loss_diag": 9.686118573881686e-11, "ewc_loss_parallel": 2.510205376893282e-10, "grad_norm": 4.98308801651001, "learning_rate": 1.1445527766002543e-08, "loss": 0.8732, "mean_token_accuracy": 0.757544219493866, "num_tokens": 1040296.0, "step": 28 }, { "epoch": 0.003689098079124793, "ewc_loss": 2.868473529815674e-07, "ewc_loss_diag": 1.0595613275654614e-10, "ewc_loss_parallel": 2.764863893389702e-10, "grad_norm": 4.421478271484375, "learning_rate": 1.1869436201780416e-08, "loss": 0.7656, "mean_token_accuracy": 0.785152792930603, "num_tokens": 1081711.0, "step": 29 }, { "epoch": 0.0038163083577153036, "ewc_loss": 3.073364496231079e-07, "ewc_loss_diag": 1.1141310096718371e-10, "ewc_loss_parallel": 2.964952727779746e-10, "grad_norm": 4.803239822387695, "learning_rate": 1.2293344637558287e-08, "loss": 0.812, "mean_token_accuracy": 0.775603175163269, "num_tokens": 1120556.0, "step": 30 }, { "epoch": 0.003943518636305814, "ewc_loss": 3.3527612686157227e-07, "ewc_loss_diag": 1.2460077414289117e-10, "ewc_loss_parallel": 3.219611244276166e-10, "grad_norm": 4.776388168334961, "learning_rate": 1.2717253073336159e-08, "loss": 0.7643, "mean_token_accuracy": 0.7894695997238159, "num_tokens": 1157723.0, "step": 31 }, { "epoch": 0.004070728914896324, "ewc_loss": 3.632158041000366e-07, "ewc_loss_diag": 1.382431946694851e-10, "ewc_loss_parallel": 3.510649548843503e-10, "grad_norm": 4.6882734298706055, "learning_rate": 1.314116150911403e-08, "loss": 0.8102, "mean_token_accuracy": 0.7743015289306641, "num_tokens": 1197879.0, "step": 32 }, { "epoch": 0.004197939193486834, "ewc_loss": 3.818422555923462e-07, "ewc_loss_diag": 1.446096575818956e-10, "ewc_loss_parallel": 3.67435859516263e-10, "grad_norm": 4.802165508270264, "learning_rate": 1.3565069944891903e-08, "loss": 0.8107, "mean_token_accuracy": 0.7752436399459839, "num_tokens": 1237342.0, "step": 33 }, { "epoch": 0.004325149472077344, "ewc_loss": 4.0605664253234863e-07, "ewc_loss_diag": 1.546140993013978e-10, "ewc_loss_parallel": 3.9108272176235914e-10, "grad_norm": 4.491925239562988, "learning_rate": 1.3988978380669775e-08, "loss": 0.7822, "mean_token_accuracy": 0.7824920415878296, "num_tokens": 1280197.0, "step": 34 }, { "epoch": 0.004452359750667854, "ewc_loss": 5.103647708892822e-07, "ewc_loss_diag": 1.90084392670542e-10, "ewc_loss_parallel": 4.911271389573812e-10, "grad_norm": 4.635449409484863, "learning_rate": 1.4412886816447646e-08, "loss": 0.7641, "mean_token_accuracy": 0.7848153114318848, "num_tokens": 1318625.0, "step": 35 }, { "epoch": 0.004579570029258364, "ewc_loss": 1.0505318641662598e-06, "ewc_loss_diag": 5.056790541857481e-10, "ewc_loss_parallel": 9.968061931431293e-10, "grad_norm": 4.790405750274658, "learning_rate": 1.4836795252225519e-08, "loss": 0.822, "mean_token_accuracy": 0.7722233533859253, "num_tokens": 1356868.0, "step": 36 }, { "epoch": 0.004706780307848874, "ewc_loss": 1.4603137969970703e-06, "ewc_loss_diag": 6.257323548197746e-10, "ewc_loss_parallel": 1.4042598195374012e-09, "grad_norm": 4.837070465087891, "learning_rate": 1.526070368800339e-08, "loss": 0.7937, "mean_token_accuracy": 0.7809069752693176, "num_tokens": 1394696.0, "step": 37 }, { "epoch": 0.004833990586439385, "ewc_loss": 1.6987323760986328e-06, "ewc_loss_diag": 7.166818249970675e-10, "ewc_loss_parallel": 1.622538547962904e-09, "grad_norm": 4.3981122970581055, "learning_rate": 1.5684612123781262e-08, "loss": 0.7827, "mean_token_accuracy": 0.7826502919197083, "num_tokens": 1438738.0, "step": 38 }, { "epoch": 0.004961200865029895, "ewc_loss": 1.817941665649414e-06, "ewc_loss_diag": 7.566995918750763e-10, "ewc_loss_parallel": 1.7462298274040222e-09, "grad_norm": 4.765013694763184, "learning_rate": 1.6108520559559135e-08, "loss": 0.748, "mean_token_accuracy": 0.7924995422363281, "num_tokens": 1475089.0, "step": 39 }, { "epoch": 0.005088411143620405, "ewc_loss": 1.9222497940063477e-06, "ewc_loss_diag": 8.076312951743603e-10, "ewc_loss_parallel": 1.8480932340025902e-09, "grad_norm": 4.731200218200684, "learning_rate": 1.6532428995337004e-08, "loss": 0.8081, "mean_token_accuracy": 0.7749139666557312, "num_tokens": 1514566.0, "step": 40 }, { "epoch": 0.005215621422210915, "ewc_loss": 2.0116567611694336e-06, "ewc_loss_diag": 8.512870408594608e-10, "ewc_loss_parallel": 1.9208528101444244e-09, "grad_norm": 4.962207794189453, "learning_rate": 1.6956337431114877e-08, "loss": 0.8233, "mean_token_accuracy": 0.7735083103179932, "num_tokens": 1552560.0, "step": 41 }, { "epoch": 0.005342831700801425, "ewc_loss": 2.041459083557129e-06, "ewc_loss_diag": 8.767528925091028e-10, "ewc_loss_parallel": 1.949956640601158e-09, "grad_norm": 5.386920928955078, "learning_rate": 1.738024586689275e-08, "loss": 0.8009, "mean_token_accuracy": 0.777316689491272, "num_tokens": 1584759.0, "step": 42 }, { "epoch": 0.005470041979391935, "ewc_loss": 2.1010637283325195e-06, "ewc_loss_diag": 9.458744898438454e-10, "ewc_loss_parallel": 2.0081643015146255e-09, "grad_norm": 4.854633331298828, "learning_rate": 1.7804154302670622e-08, "loss": 0.802, "mean_token_accuracy": 0.7756767272949219, "num_tokens": 1621825.0, "step": 43 }, { "epoch": 0.005597252257982445, "ewc_loss": 2.250075340270996e-06, "ewc_loss_diag": 1.0477378964424133e-09, "ewc_loss_parallel": 2.1391315385699272e-09, "grad_norm": 4.782246112823486, "learning_rate": 1.8228062738448494e-08, "loss": 0.8276, "mean_token_accuracy": 0.7681707143783569, "num_tokens": 1662946.0, "step": 44 }, { "epoch": 0.005724462536572955, "ewc_loss": 2.3990869522094727e-06, "ewc_loss_diag": 1.1059455573558807e-09, "ewc_loss_parallel": 2.2992026060819626e-09, "grad_norm": 4.654094696044922, "learning_rate": 1.8651971174226364e-08, "loss": 0.8155, "mean_token_accuracy": 0.7730691432952881, "num_tokens": 1699433.0, "step": 45 }, { "epoch": 0.005851672815163465, "ewc_loss": 2.562999725341797e-06, "ewc_loss_diag": 1.1568772606551647e-09, "ewc_loss_parallel": 2.444721758365631e-09, "grad_norm": 4.346685409545898, "learning_rate": 1.9075879610004236e-08, "loss": 0.7464, "mean_token_accuracy": 0.7884500026702881, "num_tokens": 1742812.0, "step": 46 }, { "epoch": 0.005978883093753975, "ewc_loss": 2.637505531311035e-06, "ewc_loss_diag": 1.2005330063402653e-09, "ewc_loss_parallel": 2.5174813345074654e-09, "grad_norm": 4.848017692565918, "learning_rate": 1.949978804578211e-08, "loss": 0.8014, "mean_token_accuracy": 0.778260350227356, "num_tokens": 1778725.0, "step": 47 }, { "epoch": 0.006106093372344485, "ewc_loss": 2.8759241104125977e-06, "ewc_loss_diag": 1.2514647096395493e-09, "ewc_loss_parallel": 2.750311978161335e-09, "grad_norm": 4.809627532958984, "learning_rate": 1.9923696481559985e-08, "loss": 0.8478, "mean_token_accuracy": 0.7614375352859497, "num_tokens": 1816592.0, "step": 48 }, { "epoch": 0.006233303650934996, "ewc_loss": 3.635883331298828e-06, "ewc_loss_diag": 1.7171259969472885e-09, "ewc_loss_parallel": 3.4633558243513107e-09, "grad_norm": 4.388747215270996, "learning_rate": 2.0347604917337857e-08, "loss": 0.74, "mean_token_accuracy": 0.7899730205535889, "num_tokens": 1859907.0, "step": 49 }, { "epoch": 0.006360513929525506, "ewc_loss": 5.900859832763672e-06, "ewc_loss_diag": 3.3614924177527428e-09, "ewc_loss_parallel": 5.558831617236137e-09, "grad_norm": 4.652761936187744, "learning_rate": 2.0771513353115727e-08, "loss": 0.7288, "mean_token_accuracy": 0.7997620701789856, "num_tokens": 1896627.0, "step": 50 }, { "epoch": 0.006487724208116016, "ewc_loss": 7.748603820800781e-06, "ewc_loss_diag": 4.336470738053322e-09, "ewc_loss_parallel": 7.30506144464016e-09, "grad_norm": 4.980456829071045, "learning_rate": 2.11954217888936e-08, "loss": 0.8447, "mean_token_accuracy": 0.7667677998542786, "num_tokens": 1934041.0, "step": 51 }, { "epoch": 0.006614934486706526, "ewc_loss": 8.881092071533203e-06, "ewc_loss_diag": 4.802132025361061e-09, "ewc_loss_parallel": 8.381903171539307e-09, "grad_norm": 4.557627201080322, "learning_rate": 2.1619330224671472e-08, "loss": 0.7734, "mean_token_accuracy": 0.7843212485313416, "num_tokens": 1976482.0, "step": 52 }, { "epoch": 0.006742144765297036, "ewc_loss": 9.775161743164062e-06, "ewc_loss_diag": 5.093170329928398e-09, "ewc_loss_parallel": 9.255018085241318e-09, "grad_norm": 5.277233123779297, "learning_rate": 2.2043238660449344e-08, "loss": 0.8029, "mean_token_accuracy": 0.7717470526695251, "num_tokens": 2009224.0, "step": 53 }, { "epoch": 0.006869355043887546, "ewc_loss": 1.0728836059570312e-05, "ewc_loss_diag": 5.3551048040390015e-09, "ewc_loss_parallel": 1.0186340659856796e-08, "grad_norm": 4.865703582763672, "learning_rate": 2.2467147096227214e-08, "loss": 0.8474, "mean_token_accuracy": 0.7617063522338867, "num_tokens": 2049235.0, "step": 54 }, { "epoch": 0.006996565322478056, "ewc_loss": 1.138448715209961e-05, "ewc_loss_diag": 5.617039278149605e-09, "ewc_loss_parallel": 1.0826624929904938e-08, "grad_norm": 4.5635175704956055, "learning_rate": 2.2891055532005086e-08, "loss": 0.8599, "mean_token_accuracy": 0.7614457607269287, "num_tokens": 2090260.0, "step": 55 }, { "epoch": 0.007123775601068566, "ewc_loss": 1.1801719665527344e-05, "ewc_loss_diag": 5.878973752260208e-09, "ewc_loss_parallel": 1.123407855629921e-08, "grad_norm": 4.999000072479248, "learning_rate": 2.331496396778296e-08, "loss": 0.8657, "mean_token_accuracy": 0.760644257068634, "num_tokens": 2126686.0, "step": 56 }, { "epoch": 0.007250985879659076, "ewc_loss": 1.245737075805664e-05, "ewc_loss_diag": 6.05359673500061e-09, "ewc_loss_parallel": 1.1874362826347351e-08, "grad_norm": 4.362381458282471, "learning_rate": 2.373887240356083e-08, "loss": 0.7675, "mean_token_accuracy": 0.7849968075752258, "num_tokens": 2171355.0, "step": 57 }, { "epoch": 0.007378196158249586, "ewc_loss": 1.2576580047607422e-05, "ewc_loss_diag": 6.28642737865448e-09, "ewc_loss_parallel": 1.1932570487260818e-08, "grad_norm": 4.5879597663879395, "learning_rate": 2.4162780839338704e-08, "loss": 0.7406, "mean_token_accuracy": 0.7930310964584351, "num_tokens": 2211660.0, "step": 58 }, { "epoch": 0.007505406436840096, "ewc_loss": 1.2814998626708984e-05, "ewc_loss_diag": 6.4319465309381485e-09, "ewc_loss_parallel": 1.2165401130914688e-08, "grad_norm": 5.534573078155518, "learning_rate": 2.4586689275116573e-08, "loss": 0.8709, "mean_token_accuracy": 0.7590556740760803, "num_tokens": 2244411.0, "step": 59 }, { "epoch": 0.007632616715430607, "ewc_loss": 1.2993812561035156e-05, "ewc_loss_diag": 6.868503987789154e-09, "ewc_loss_parallel": 1.234002411365509e-08, "grad_norm": 4.3568115234375, "learning_rate": 2.5010597710894446e-08, "loss": 0.7974, "mean_token_accuracy": 0.7708356380462646, "num_tokens": 2285253.0, "step": 60 }, { "epoch": 0.007759826994021117, "ewc_loss": 1.33514404296875e-05, "ewc_loss_diag": 7.450580596923828e-09, "ewc_loss_parallel": 1.257285475730896e-08, "grad_norm": 4.483847141265869, "learning_rate": 2.5434506146672318e-08, "loss": 0.762, "mean_token_accuracy": 0.7854892015457153, "num_tokens": 2328577.0, "step": 61 }, { "epoch": 0.007887037272611627, "ewc_loss": 1.3589859008789062e-05, "ewc_loss_diag": 7.8580342233181e-09, "ewc_loss_parallel": 1.280568540096283e-08, "grad_norm": 4.7122883796691895, "learning_rate": 2.585841458245019e-08, "loss": 0.842, "mean_token_accuracy": 0.7701469659805298, "num_tokens": 2366524.0, "step": 62 }, { "epoch": 0.008014247551202136, "ewc_loss": 1.4126300811767578e-05, "ewc_loss_diag": 8.149072527885437e-09, "ewc_loss_parallel": 1.3271346688270569e-08, "grad_norm": 4.205564498901367, "learning_rate": 2.628232301822806e-08, "loss": 0.7557, "mean_token_accuracy": 0.7897399663925171, "num_tokens": 2408628.0, "step": 63 }, { "epoch": 0.008141457829792647, "ewc_loss": 1.430511474609375e-05, "ewc_loss_diag": 8.381903171539307e-09, "ewc_loss_parallel": 1.3445969671010971e-08, "grad_norm": 4.2738823890686035, "learning_rate": 2.6706231454005933e-08, "loss": 0.795, "mean_token_accuracy": 0.7754863500595093, "num_tokens": 2451800.0, "step": 64 }, { "epoch": 0.008268668108383158, "ewc_loss": 1.4603137969970703e-05, "ewc_loss_diag": 8.731149137020111e-09, "ewc_loss_parallel": 1.3737007975578308e-08, "grad_norm": 5.674116611480713, "learning_rate": 2.7130139889783805e-08, "loss": 0.8661, "mean_token_accuracy": 0.7548186779022217, "num_tokens": 2481448.0, "step": 65 }, { "epoch": 0.008395878386973667, "ewc_loss": 1.5139579772949219e-05, "ewc_loss_diag": 9.022187441587448e-09, "ewc_loss_parallel": 1.4260876923799515e-08, "grad_norm": 4.268650531768799, "learning_rate": 2.7554048325561678e-08, "loss": 0.7929, "mean_token_accuracy": 0.7756462097167969, "num_tokens": 2526339.0, "step": 66 }, { "epoch": 0.008523088665564178, "ewc_loss": 1.5735626220703125e-05, "ewc_loss_diag": 9.255018085241318e-09, "ewc_loss_parallel": 1.4842953532934189e-08, "grad_norm": 4.427706241607666, "learning_rate": 2.797795676133955e-08, "loss": 0.8177, "mean_token_accuracy": 0.7703561782836914, "num_tokens": 2570691.0, "step": 67 }, { "epoch": 0.008650298944154687, "ewc_loss": 1.6450881958007812e-05, "ewc_loss_diag": 9.89530235528946e-09, "ewc_loss_parallel": 1.548323780298233e-08, "grad_norm": 4.576381683349609, "learning_rate": 2.840186519711742e-08, "loss": 0.8044, "mean_token_accuracy": 0.7749950885772705, "num_tokens": 2609207.0, "step": 68 }, { "epoch": 0.008777509222745198, "ewc_loss": 2.2649765014648438e-05, "ewc_loss_diag": 1.4319084584712982e-08, "ewc_loss_parallel": 2.1187588572502136e-08, "grad_norm": 4.900610446929932, "learning_rate": 2.8825773632895292e-08, "loss": 0.7928, "mean_token_accuracy": 0.7774732112884521, "num_tokens": 2645494.0, "step": 69 }, { "epoch": 0.008904719501335707, "ewc_loss": 3.361701965332031e-05, "ewc_loss_diag": 2.0489096641540527e-08, "ewc_loss_parallel": 3.14321368932724e-08, "grad_norm": 4.444159984588623, "learning_rate": 2.9249682068673165e-08, "loss": 0.7498, "mean_token_accuracy": 0.7913873195648193, "num_tokens": 2686258.0, "step": 70 }, { "epoch": 0.009031929779926218, "ewc_loss": 4.124641418457031e-05, "ewc_loss_diag": 2.514570951461792e-08, "ewc_loss_parallel": 3.864988684654236e-08, "grad_norm": 5.0149006843566895, "learning_rate": 2.9673590504451037e-08, "loss": 0.8044, "mean_token_accuracy": 0.7778772115707397, "num_tokens": 2720544.0, "step": 71 }, { "epoch": 0.009159140058516728, "ewc_loss": 4.76837158203125e-05, "ewc_loss_diag": 2.7939677238464355e-08, "ewc_loss_parallel": 4.470348358154297e-08, "grad_norm": 4.821709156036377, "learning_rate": 3.0097498940228907e-08, "loss": 0.7677, "mean_token_accuracy": 0.7854229807853699, "num_tokens": 2758068.0, "step": 72 }, { "epoch": 0.009286350337107238, "ewc_loss": 5.2928924560546875e-05, "ewc_loss_diag": 2.9685907065868378e-08, "ewc_loss_parallel": 5.005858838558197e-08, "grad_norm": 5.043861389160156, "learning_rate": 3.052140737600678e-08, "loss": 0.8517, "mean_token_accuracy": 0.7631268501281738, "num_tokens": 2793342.0, "step": 73 }, { "epoch": 0.009413560615697748, "ewc_loss": 5.626678466796875e-05, "ewc_loss_diag": 3.073364496231079e-08, "ewc_loss_parallel": 5.3085386753082275e-08, "grad_norm": 5.207295894622803, "learning_rate": 3.094531581178465e-08, "loss": 0.8886, "mean_token_accuracy": 0.755763053894043, "num_tokens": 2828003.0, "step": 74 }, { "epoch": 0.009540770894288259, "ewc_loss": 5.936622619628906e-05, "ewc_loss_diag": 3.189779818058014e-08, "ewc_loss_parallel": 5.634501576423645e-08, "grad_norm": 4.269049167633057, "learning_rate": 3.1369224247562524e-08, "loss": 0.7639, "mean_token_accuracy": 0.7843363285064697, "num_tokens": 2874755.0, "step": 75 }, { "epoch": 0.00966798117287877, "ewc_loss": 6.031990051269531e-05, "ewc_loss_diag": 3.306195139884949e-08, "ewc_loss_parallel": 5.704350769519806e-08, "grad_norm": 4.622293472290039, "learning_rate": 3.17931326833404e-08, "loss": 0.786, "mean_token_accuracy": 0.7767789959907532, "num_tokens": 2909703.0, "step": 76 }, { "epoch": 0.009795191451469279, "ewc_loss": 6.079673767089844e-05, "ewc_loss_diag": 3.4226104617118835e-08, "ewc_loss_parallel": 5.75091689825058e-08, "grad_norm": 4.151971817016602, "learning_rate": 3.221704111911827e-08, "loss": 0.7162, "mean_token_accuracy": 0.7989228963851929, "num_tokens": 2954020.0, "step": 77 }, { "epoch": 0.00992240173005979, "ewc_loss": 6.079673767089844e-05, "ewc_loss_diag": 3.562308847904205e-08, "ewc_loss_parallel": 5.727633833885193e-08, "grad_norm": 5.212409496307373, "learning_rate": 3.264094955489614e-08, "loss": 0.8116, "mean_token_accuracy": 0.7750231027603149, "num_tokens": 2984894.0, "step": 78 }, { "epoch": 0.010049612008650299, "ewc_loss": 6.103515625e-05, "ewc_loss_diag": 3.748573362827301e-08, "ewc_loss_parallel": 5.727633833885193e-08, "grad_norm": 4.939089775085449, "learning_rate": 3.306485799067401e-08, "loss": 0.7184, "mean_token_accuracy": 0.795691728591919, "num_tokens": 3017773.0, "step": 79 }, { "epoch": 0.01017682228724081, "ewc_loss": 6.151199340820312e-05, "ewc_loss_diag": 3.864988684654236e-08, "ewc_loss_parallel": 5.774199962615967e-08, "grad_norm": 4.277116298675537, "learning_rate": 3.348876642645188e-08, "loss": 0.7179, "mean_token_accuracy": 0.7983066439628601, "num_tokens": 3059739.0, "step": 80 }, { "epoch": 0.010304032565831319, "ewc_loss": 6.198883056640625e-05, "ewc_loss_diag": 3.91155481338501e-08, "ewc_loss_parallel": 5.797483026981354e-08, "grad_norm": 4.857213020324707, "learning_rate": 3.391267486222975e-08, "loss": 0.8098, "mean_token_accuracy": 0.7716346383094788, "num_tokens": 3094910.0, "step": 81 }, { "epoch": 0.01043124284442183, "ewc_loss": 6.151199340820312e-05, "ewc_loss_diag": 3.958120942115784e-08, "ewc_loss_parallel": 5.75091689825058e-08, "grad_norm": 4.568922996520996, "learning_rate": 3.4336583298007626e-08, "loss": 0.8138, "mean_token_accuracy": 0.77337646484375, "num_tokens": 3133884.0, "step": 82 }, { "epoch": 0.010558453123012339, "ewc_loss": 6.079673767089844e-05, "ewc_loss_diag": 4.0512531995773315e-08, "ewc_loss_parallel": 5.681067705154419e-08, "grad_norm": 5.157546520233154, "learning_rate": 3.47604917337855e-08, "loss": 0.8461, "mean_token_accuracy": 0.7603965997695923, "num_tokens": 3166755.0, "step": 83 }, { "epoch": 0.01068566340160285, "ewc_loss": 6.008148193359375e-05, "ewc_loss_diag": 4.1676685214042664e-08, "ewc_loss_parallel": 5.611218512058258e-08, "grad_norm": 4.706188678741455, "learning_rate": 3.518440016956337e-08, "loss": 0.8146, "mean_token_accuracy": 0.7712476253509521, "num_tokens": 3204825.0, "step": 84 }, { "epoch": 0.010812873680193359, "ewc_loss": 5.984306335449219e-05, "ewc_loss_diag": 4.493631422519684e-08, "ewc_loss_parallel": 5.541369318962097e-08, "grad_norm": 4.7194976806640625, "learning_rate": 3.5608308605341244e-08, "loss": 0.8505, "mean_token_accuracy": 0.7628300189971924, "num_tokens": 3244036.0, "step": 85 }, { "epoch": 0.01094008395878387, "ewc_loss": 5.936622619628906e-05, "ewc_loss_diag": 4.7963112592697144e-08, "ewc_loss_parallel": 5.448237061500549e-08, "grad_norm": 4.481651306152344, "learning_rate": 3.6032217041119116e-08, "loss": 0.8379, "mean_token_accuracy": 0.7666102051734924, "num_tokens": 3285969.0, "step": 86 }, { "epoch": 0.01106729423737438, "ewc_loss": 5.9604644775390625e-05, "ewc_loss_diag": 5.029141902923584e-08, "ewc_loss_parallel": 5.448237061500549e-08, "grad_norm": 4.443980693817139, "learning_rate": 3.645612547689699e-08, "loss": 0.8088, "mean_token_accuracy": 0.7752890586853027, "num_tokens": 3327648.0, "step": 87 }, { "epoch": 0.01119450451596489, "ewc_loss": 6.008148193359375e-05, "ewc_loss_diag": 5.192123353481293e-08, "ewc_loss_parallel": 5.494803190231323e-08, "grad_norm": 4.595998764038086, "learning_rate": 3.6880033912674855e-08, "loss": 0.8121, "mean_token_accuracy": 0.7726224660873413, "num_tokens": 3367399.0, "step": 88 }, { "epoch": 0.0113217147945554, "ewc_loss": 6.008148193359375e-05, "ewc_loss_diag": 5.3318217396736145e-08, "ewc_loss_parallel": 5.494803190231323e-08, "grad_norm": 4.787242412567139, "learning_rate": 3.730394234845273e-08, "loss": 0.7952, "mean_token_accuracy": 0.7798758149147034, "num_tokens": 3405402.0, "step": 89 }, { "epoch": 0.01144892507314591, "ewc_loss": 6.103515625e-05, "ewc_loss_diag": 5.471520125865936e-08, "ewc_loss_parallel": 5.587935447692871e-08, "grad_norm": 4.884618759155273, "learning_rate": 3.77278507842306e-08, "loss": 0.8788, "mean_token_accuracy": 0.7541158199310303, "num_tokens": 3441791.0, "step": 90 }, { "epoch": 0.01157613535173642, "ewc_loss": 6.29425048828125e-05, "ewc_loss_diag": 5.657784640789032e-08, "ewc_loss_parallel": 5.75091689825058e-08, "grad_norm": 4.83642053604126, "learning_rate": 3.815175922000847e-08, "loss": 0.8383, "mean_token_accuracy": 0.7690328359603882, "num_tokens": 3480151.0, "step": 91 }, { "epoch": 0.01170334563032693, "ewc_loss": 6.437301635742188e-05, "ewc_loss_diag": 5.820766091346741e-08, "ewc_loss_parallel": 5.8673322200775146e-08, "grad_norm": 4.728335380554199, "learning_rate": 3.8575667655786345e-08, "loss": 0.8017, "mean_token_accuracy": 0.7760720252990723, "num_tokens": 3516867.0, "step": 92 }, { "epoch": 0.01183055590891744, "ewc_loss": 6.67572021484375e-05, "ewc_loss_diag": 5.960464477539063e-08, "ewc_loss_parallel": 6.100162863731384e-08, "grad_norm": 4.310719966888428, "learning_rate": 3.899957609156422e-08, "loss": 0.7779, "mean_token_accuracy": 0.7828134894371033, "num_tokens": 3557466.0, "step": 93 }, { "epoch": 0.01195776618750795, "ewc_loss": 7.05718994140625e-05, "ewc_loss_diag": 6.007030606269836e-08, "ewc_loss_parallel": 6.426125764846802e-08, "grad_norm": 4.538886070251465, "learning_rate": 3.94234845273421e-08, "loss": 0.8071, "mean_token_accuracy": 0.7731052041053772, "num_tokens": 3596009.0, "step": 94 }, { "epoch": 0.012084976466098461, "ewc_loss": 8.344650268554688e-05, "ewc_loss_diag": 6.146728992462158e-08, "ewc_loss_parallel": 7.729977369308472e-08, "grad_norm": 4.746531009674072, "learning_rate": 3.984739296311997e-08, "loss": 0.7992, "mean_token_accuracy": 0.77455073595047, "num_tokens": 3631582.0, "step": 95 }, { "epoch": 0.01221218674468897, "ewc_loss": 0.00010967254638671875, "ewc_loss_diag": 6.565824151039124e-08, "ewc_loss_parallel": 1.0291114449501038e-07, "grad_norm": 4.704276084899902, "learning_rate": 4.027130139889784e-08, "loss": 0.8258, "mean_token_accuracy": 0.7678328156471252, "num_tokens": 3672855.0, "step": 96 }, { "epoch": 0.012339397023279481, "ewc_loss": 0.00015926361083984375, "ewc_loss_diag": 7.59027898311615e-08, "ewc_loss_parallel": 1.51805579662323e-07, "grad_norm": 4.518270492553711, "learning_rate": 4.0695209834675715e-08, "loss": 0.8175, "mean_token_accuracy": 0.7731975317001343, "num_tokens": 3714100.0, "step": 97 }, { "epoch": 0.012466607301869992, "ewc_loss": 0.0002117156982421875, "ewc_loss_diag": 9.033828973770142e-08, "ewc_loss_parallel": 2.0302832126617432e-07, "grad_norm": 4.61027193069458, "learning_rate": 4.111911827045358e-08, "loss": 0.7687, "mean_token_accuracy": 0.7850976586341858, "num_tokens": 3752834.0, "step": 98 }, { "epoch": 0.012593817580460501, "ewc_loss": 0.0002498626708984375, "ewc_loss_diag": 1.0244548320770264e-07, "ewc_loss_parallel": 2.4028122425079346e-07, "grad_norm": 4.662015914916992, "learning_rate": 4.154302670623145e-08, "loss": 0.7789, "mean_token_accuracy": 0.7813397645950317, "num_tokens": 3791513.0, "step": 99 }, { "epoch": 0.012721027859051012, "ewc_loss": 0.0002651214599609375, "ewc_loss_diag": 1.0896474123001099e-07, "ewc_loss_parallel": 2.5331974029541016e-07, "grad_norm": 5.144886493682861, "learning_rate": 4.1966935142009326e-08, "loss": 0.8019, "mean_token_accuracy": 0.7748047709465027, "num_tokens": 3825085.0, "step": 100 }, { "epoch": 0.012848238137641521, "ewc_loss": 0.00026702880859375, "ewc_loss_diag": 1.1641532182693481e-07, "ewc_loss_parallel": 2.551823854446411e-07, "grad_norm": 4.540340423583984, "learning_rate": 4.23908435777872e-08, "loss": 0.8019, "mean_token_accuracy": 0.7766343355178833, "num_tokens": 3866492.0, "step": 101 }, { "epoch": 0.012975448416232032, "ewc_loss": 0.0002613067626953125, "ewc_loss_diag": 1.210719347000122e-07, "ewc_loss_parallel": 2.4959444999694824e-07, "grad_norm": 4.902456283569336, "learning_rate": 4.281475201356507e-08, "loss": 0.8291, "mean_token_accuracy": 0.7635929584503174, "num_tokens": 3904053.0, "step": 102 }, { "epoch": 0.013102658694822541, "ewc_loss": 0.0002613067626953125, "ewc_loss_diag": 1.2665987014770508e-07, "ewc_loss_parallel": 2.477318048477173e-07, "grad_norm": 4.899242877960205, "learning_rate": 4.3238660449342943e-08, "loss": 0.7377, "mean_token_accuracy": 0.7951051592826843, "num_tokens": 3935118.0, "step": 103 }, { "epoch": 0.013229868973413052, "ewc_loss": 0.0002593994140625, "ewc_loss_diag": 1.3131648302078247e-07, "ewc_loss_parallel": 2.4586915969848633e-07, "grad_norm": 4.424589157104492, "learning_rate": 4.3662568885120816e-08, "loss": 0.7878, "mean_token_accuracy": 0.7764580249786377, "num_tokens": 3977899.0, "step": 104 }, { "epoch": 0.013357079252003561, "ewc_loss": 0.0002574920654296875, "ewc_loss_diag": 1.3504177331924438e-07, "ewc_loss_parallel": 2.4400651454925537e-07, "grad_norm": 4.442825794219971, "learning_rate": 4.408647732089869e-08, "loss": 0.8081, "mean_token_accuracy": 0.7735345959663391, "num_tokens": 4019416.0, "step": 105 }, { "epoch": 0.013484289530594072, "ewc_loss": 0.0002574920654296875, "ewc_loss_diag": 1.387670636177063e-07, "ewc_loss_parallel": 2.4400651454925537e-07, "grad_norm": 4.353634834289551, "learning_rate": 4.451038575667656e-08, "loss": 0.797, "mean_token_accuracy": 0.7733345031738281, "num_tokens": 4063887.0, "step": 106 }, { "epoch": 0.013611499809184581, "ewc_loss": 0.00025177001953125, "ewc_loss_diag": 1.424923539161682e-07, "ewc_loss_parallel": 2.384185791015625e-07, "grad_norm": 4.505864143371582, "learning_rate": 4.493429419245443e-08, "loss": 0.8281, "mean_token_accuracy": 0.7677685022354126, "num_tokens": 4101309.0, "step": 107 }, { "epoch": 0.013738710087775092, "ewc_loss": 0.0002498626708984375, "ewc_loss_diag": 1.4621764421463013e-07, "ewc_loss_parallel": 2.3469328880310059e-07, "grad_norm": 4.640070915222168, "learning_rate": 4.53582026282323e-08, "loss": 0.8123, "mean_token_accuracy": 0.7703113555908203, "num_tokens": 4136669.0, "step": 108 }, { "epoch": 0.013865920366365603, "ewc_loss": 0.0002460479736328125, "ewc_loss_diag": 1.4901161193847656e-07, "ewc_loss_parallel": 2.3096799850463867e-07, "grad_norm": 4.1892266273498535, "learning_rate": 4.578211106401017e-08, "loss": 0.7432, "mean_token_accuracy": 0.7901679277420044, "num_tokens": 4178583.0, "step": 109 }, { "epoch": 0.013993130644956112, "ewc_loss": 0.0002422332763671875, "ewc_loss_diag": 1.5273690223693848e-07, "ewc_loss_parallel": 2.2724270820617676e-07, "grad_norm": 4.194305419921875, "learning_rate": 4.6206019499788045e-08, "loss": 0.7736, "mean_token_accuracy": 0.7775267362594604, "num_tokens": 4219552.0, "step": 110 }, { "epoch": 0.014120340923546623, "ewc_loss": 0.00023555755615234375, "ewc_loss_diag": 1.555308699607849e-07, "ewc_loss_parallel": 2.207234501838684e-07, "grad_norm": 4.236396789550781, "learning_rate": 4.662992793556592e-08, "loss": 0.7665, "mean_token_accuracy": 0.7851260900497437, "num_tokens": 4264487.0, "step": 111 }, { "epoch": 0.014247551202137132, "ewc_loss": 0.00022983551025390625, "ewc_loss_diag": 1.5925616025924683e-07, "ewc_loss_parallel": 2.1327286958694458e-07, "grad_norm": 4.336306571960449, "learning_rate": 4.705383637134379e-08, "loss": 0.7329, "mean_token_accuracy": 0.7902770638465881, "num_tokens": 4303015.0, "step": 112 }, { "epoch": 0.014374761480727643, "ewc_loss": 0.000225067138671875, "ewc_loss_diag": 1.6391277313232422e-07, "ewc_loss_parallel": 2.086162567138672e-07, "grad_norm": 4.276446342468262, "learning_rate": 4.747774480712166e-08, "loss": 0.7877, "mean_token_accuracy": 0.7769608497619629, "num_tokens": 4345446.0, "step": 113 }, { "epoch": 0.014501971759318152, "ewc_loss": 0.00022029876708984375, "ewc_loss_diag": 1.6763806343078613e-07, "ewc_loss_parallel": 2.0302832126617432e-07, "grad_norm": 4.388430595397949, "learning_rate": 4.7901653242899535e-08, "loss": 0.7398, "mean_token_accuracy": 0.7908351421356201, "num_tokens": 4383854.0, "step": 114 }, { "epoch": 0.014629182037908663, "ewc_loss": 0.00021457672119140625, "ewc_loss_diag": 1.695007085800171e-07, "ewc_loss_parallel": 1.9744038581848145e-07, "grad_norm": 4.452174663543701, "learning_rate": 4.832556167867741e-08, "loss": 0.8088, "mean_token_accuracy": 0.773084282875061, "num_tokens": 4420750.0, "step": 115 }, { "epoch": 0.014756392316499172, "ewc_loss": 0.0002117156982421875, "ewc_loss_diag": 1.7415732145309448e-07, "ewc_loss_parallel": 1.94646418094635e-07, "grad_norm": 4.584152698516846, "learning_rate": 4.8749470114455274e-08, "loss": 0.8059, "mean_token_accuracy": 0.7732776403427124, "num_tokens": 4458761.0, "step": 116 }, { "epoch": 0.014883602595089683, "ewc_loss": 0.00020599365234375, "ewc_loss_diag": 1.7695128917694092e-07, "ewc_loss_parallel": 1.8812716007232666e-07, "grad_norm": 4.360849857330322, "learning_rate": 4.9173378550233146e-08, "loss": 0.766, "mean_token_accuracy": 0.7809910774230957, "num_tokens": 4496547.0, "step": 117 }, { "epoch": 0.015010812873680193, "ewc_loss": 0.0002002716064453125, "ewc_loss_diag": 1.816079020500183e-07, "ewc_loss_parallel": 1.825392246246338e-07, "grad_norm": 4.394128322601318, "learning_rate": 4.959728698601102e-08, "loss": 0.7639, "mean_token_accuracy": 0.790017306804657, "num_tokens": 4533357.0, "step": 118 }, { "epoch": 0.015138023152270703, "ewc_loss": 0.0001983642578125, "ewc_loss_diag": 1.8719583749771118e-07, "ewc_loss_parallel": 1.7974525690078735e-07, "grad_norm": 4.2169508934021, "learning_rate": 5.002119542178889e-08, "loss": 0.7384, "mean_token_accuracy": 0.7906313538551331, "num_tokens": 4573570.0, "step": 119 }, { "epoch": 0.015265233430861214, "ewc_loss": 0.0001964569091796875, "ewc_loss_diag": 1.9371509552001953e-07, "ewc_loss_parallel": 1.7695128917694092e-07, "grad_norm": 4.410034656524658, "learning_rate": 5.0445103857566764e-08, "loss": 0.7934, "mean_token_accuracy": 0.7751603126525879, "num_tokens": 4613195.0, "step": 120 }, { "epoch": 0.015392443709451724, "ewc_loss": 0.0001964569091796875, "ewc_loss_diag": 2.0209699869155884e-07, "ewc_loss_parallel": 1.7601996660232544e-07, "grad_norm": 4.064315319061279, "learning_rate": 5.0869012293344637e-08, "loss": 0.7139, "mean_token_accuracy": 0.7978008389472961, "num_tokens": 4657230.0, "step": 121 }, { "epoch": 0.015519653988042234, "ewc_loss": 0.000194549560546875, "ewc_loss_diag": 2.0954757928848267e-07, "ewc_loss_parallel": 1.73225998878479e-07, "grad_norm": 4.224815368652344, "learning_rate": 5.129292072912251e-08, "loss": 0.7878, "mean_token_accuracy": 0.7806862592697144, "num_tokens": 4701436.0, "step": 122 }, { "epoch": 0.015646864266632744, "ewc_loss": 0.0001926422119140625, "ewc_loss_diag": 2.1792948246002197e-07, "ewc_loss_parallel": 1.7043203115463257e-07, "grad_norm": 4.327423572540283, "learning_rate": 5.171682916490038e-08, "loss": 0.7753, "mean_token_accuracy": 0.7806490659713745, "num_tokens": 4741806.0, "step": 123 }, { "epoch": 0.015774074545223254, "ewc_loss": 0.00019168853759765625, "ewc_loss_diag": 2.2631138563156128e-07, "ewc_loss_parallel": 1.685693860054016e-07, "grad_norm": 4.520432472229004, "learning_rate": 5.2140737600678254e-08, "loss": 0.8127, "mean_token_accuracy": 0.7656079530715942, "num_tokens": 4780671.0, "step": 124 }, { "epoch": 0.015901284823813765, "ewc_loss": 0.00019168853759765625, "ewc_loss_diag": 2.3469328880310059e-07, "ewc_loss_parallel": 1.6763806343078613e-07, "grad_norm": 4.512225151062012, "learning_rate": 5.256464603645612e-08, "loss": 0.7611, "mean_token_accuracy": 0.7863945960998535, "num_tokens": 4816879.0, "step": 125 }, { "epoch": 0.016028495102404273, "ewc_loss": 0.00019168853759765625, "ewc_loss_diag": 2.384185791015625e-07, "ewc_loss_parallel": 1.6763806343078613e-07, "grad_norm": 4.385119915008545, "learning_rate": 5.298855447223399e-08, "loss": 0.8134, "mean_token_accuracy": 0.767613410949707, "num_tokens": 4858704.0, "step": 126 }, { "epoch": 0.016155705380994784, "ewc_loss": 0.00019359588623046875, "ewc_loss_diag": 2.4400651454925537e-07, "ewc_loss_parallel": 1.685693860054016e-07, "grad_norm": 4.354907989501953, "learning_rate": 5.3412462908011865e-08, "loss": 0.7707, "mean_token_accuracy": 0.7827750444412231, "num_tokens": 4896780.0, "step": 127 }, { "epoch": 0.016282915659585295, "ewc_loss": 0.00019550323486328125, "ewc_loss_diag": 2.514570951461792e-07, "ewc_loss_parallel": 1.7043203115463257e-07, "grad_norm": 4.579240798950195, "learning_rate": 5.383637134378974e-08, "loss": 0.8005, "mean_token_accuracy": 0.7721003293991089, "num_tokens": 4932108.0, "step": 128 }, { "epoch": 0.016410125938175806, "ewc_loss": 0.00020503997802734375, "ewc_loss_diag": 2.5704503059387207e-07, "ewc_loss_parallel": 1.7881393432617188e-07, "grad_norm": 4.311951160430908, "learning_rate": 5.426027977956761e-08, "loss": 0.7767, "mean_token_accuracy": 0.7825612425804138, "num_tokens": 4970384.0, "step": 129 }, { "epoch": 0.016537336216766316, "ewc_loss": 0.00021457672119140625, "ewc_loss_diag": 2.682209014892578e-07, "ewc_loss_parallel": 1.8812716007232666e-07, "grad_norm": 4.632563591003418, "learning_rate": 5.468418821534548e-08, "loss": 0.7852, "mean_token_accuracy": 0.7773352265357971, "num_tokens": 5005996.0, "step": 130 }, { "epoch": 0.016664546495356824, "ewc_loss": 0.00023555755615234375, "ewc_loss_diag": 2.738088369369507e-07, "ewc_loss_parallel": 2.076849341392517e-07, "grad_norm": 4.446529388427734, "learning_rate": 5.5108096651123356e-08, "loss": 0.765, "mean_token_accuracy": 0.7817535400390625, "num_tokens": 5044208.0, "step": 131 }, { "epoch": 0.016791756773947335, "ewc_loss": 0.0002613067626953125, "ewc_loss_diag": 2.849847078323364e-07, "ewc_loss_parallel": 2.337619662284851e-07, "grad_norm": 4.347762107849121, "learning_rate": 5.553200508690123e-08, "loss": 0.7798, "mean_token_accuracy": 0.7762179374694824, "num_tokens": 5085173.0, "step": 132 }, { "epoch": 0.016918967052537846, "ewc_loss": 0.00030517578125, "ewc_loss_diag": 2.9243528842926025e-07, "ewc_loss_parallel": 2.7567148208618164e-07, "grad_norm": 4.808290004730225, "learning_rate": 5.59559135226791e-08, "loss": 0.7959, "mean_token_accuracy": 0.7731851935386658, "num_tokens": 5118440.0, "step": 133 }, { "epoch": 0.017046177331128357, "ewc_loss": 0.0003643035888671875, "ewc_loss_diag": 2.998858690261841e-07, "ewc_loss_parallel": 3.334134817123413e-07, "grad_norm": 4.885855197906494, "learning_rate": 5.637982195845697e-08, "loss": 0.8001, "mean_token_accuracy": 0.774315595626831, "num_tokens": 5155202.0, "step": 134 }, { "epoch": 0.017173387609718864, "ewc_loss": 0.00045013427734375, "ewc_loss_diag": 3.129243850708008e-07, "ewc_loss_parallel": 4.1909515857696533e-07, "grad_norm": 4.762484550476074, "learning_rate": 5.680373039423484e-08, "loss": 0.7855, "mean_token_accuracy": 0.7795648574829102, "num_tokens": 5193240.0, "step": 135 }, { "epoch": 0.017300597888309375, "ewc_loss": 0.00055694580078125, "ewc_loss_diag": 3.2782554626464844e-07, "ewc_loss_parallel": 5.21540641784668e-07, "grad_norm": 4.831378936767578, "learning_rate": 5.722763883001271e-08, "loss": 0.773, "mean_token_accuracy": 0.7809896469116211, "num_tokens": 5230975.0, "step": 136 }, { "epoch": 0.017427808166899886, "ewc_loss": 0.00063323974609375, "ewc_loss_diag": 3.4831464290618896e-07, "ewc_loss_parallel": 5.997717380523682e-07, "grad_norm": 5.327452659606934, "learning_rate": 5.7651547265790585e-08, "loss": 0.8348, "mean_token_accuracy": 0.7640285491943359, "num_tokens": 5264786.0, "step": 137 }, { "epoch": 0.017555018445490397, "ewc_loss": 0.000720977783203125, "ewc_loss_diag": 3.594905138015747e-07, "ewc_loss_parallel": 6.854534149169922e-07, "grad_norm": 4.902443885803223, "learning_rate": 5.807545570156846e-08, "loss": 0.7969, "mean_token_accuracy": 0.7740215063095093, "num_tokens": 5301554.0, "step": 138 }, { "epoch": 0.017682228724080904, "ewc_loss": 0.000766754150390625, "ewc_loss_diag": 3.725290298461914e-07, "ewc_loss_parallel": 7.301568984985352e-07, "grad_norm": 4.776069641113281, "learning_rate": 5.849936413734633e-08, "loss": 0.7355, "mean_token_accuracy": 0.7902962565422058, "num_tokens": 5337356.0, "step": 139 }, { "epoch": 0.017809439002671415, "ewc_loss": 0.00079345703125, "ewc_loss_diag": 3.818422555923462e-07, "ewc_loss_parallel": 7.562339305877686e-07, "grad_norm": 4.6341423988342285, "learning_rate": 5.89232725731242e-08, "loss": 0.8067, "mean_token_accuracy": 0.7738673090934753, "num_tokens": 5375767.0, "step": 140 }, { "epoch": 0.017936649281261926, "ewc_loss": 0.000820159912109375, "ewc_loss_diag": 4.023313522338867e-07, "ewc_loss_parallel": 7.7858567237854e-07, "grad_norm": 4.5831098556518555, "learning_rate": 5.9347181008902075e-08, "loss": 0.7574, "mean_token_accuracy": 0.7828431129455566, "num_tokens": 5413683.0, "step": 141 }, { "epoch": 0.018063859559852437, "ewc_loss": 0.000827789306640625, "ewc_loss_diag": 4.209578037261963e-07, "ewc_loss_parallel": 7.860362529754639e-07, "grad_norm": 4.283658504486084, "learning_rate": 5.977108944467995e-08, "loss": 0.7383, "mean_token_accuracy": 0.7884490489959717, "num_tokens": 5457750.0, "step": 142 }, { "epoch": 0.018191069838442948, "ewc_loss": 0.0008392333984375, "ewc_loss_diag": 4.4517219066619873e-07, "ewc_loss_parallel": 7.934868335723877e-07, "grad_norm": 4.5780839920043945, "learning_rate": 6.019499788045781e-08, "loss": 0.7333, "mean_token_accuracy": 0.7899456024169922, "num_tokens": 5495743.0, "step": 143 }, { "epoch": 0.018318280117033455, "ewc_loss": 0.00083160400390625, "ewc_loss_diag": 4.5634806156158447e-07, "ewc_loss_parallel": 7.860362529754639e-07, "grad_norm": 4.601631164550781, "learning_rate": 6.061890631623569e-08, "loss": 0.7464, "mean_token_accuracy": 0.7818400859832764, "num_tokens": 5532198.0, "step": 144 }, { "epoch": 0.018445490395623966, "ewc_loss": 0.000823974609375, "ewc_loss_diag": 4.6566128730773926e-07, "ewc_loss_parallel": 7.7858567237854e-07, "grad_norm": 4.7502121925354, "learning_rate": 6.104281475201356e-08, "loss": 0.7588, "mean_token_accuracy": 0.7827471494674683, "num_tokens": 5568977.0, "step": 145 }, { "epoch": 0.018572700674214477, "ewc_loss": 0.00083160400390625, "ewc_loss_diag": 4.76837158203125e-07, "ewc_loss_parallel": 7.860362529754639e-07, "grad_norm": 4.6565141677856445, "learning_rate": 6.146672318779143e-08, "loss": 0.7711, "mean_token_accuracy": 0.7737069129943848, "num_tokens": 5606229.0, "step": 146 }, { "epoch": 0.018699910952804988, "ewc_loss": 0.00083160400390625, "ewc_loss_diag": 4.842877388000488e-07, "ewc_loss_parallel": 7.82310962677002e-07, "grad_norm": 4.410985946655273, "learning_rate": 6.18906316235693e-08, "loss": 0.7724, "mean_token_accuracy": 0.7763077020645142, "num_tokens": 5649828.0, "step": 147 }, { "epoch": 0.018827121231395495, "ewc_loss": 0.000820159912109375, "ewc_loss_diag": 4.917383193969727e-07, "ewc_loss_parallel": 7.711350917816162e-07, "grad_norm": 4.4859843254089355, "learning_rate": 6.231454005934718e-08, "loss": 0.7185, "mean_token_accuracy": 0.7965364456176758, "num_tokens": 5687433.0, "step": 148 }, { "epoch": 0.018954331509986006, "ewc_loss": 0.0008087158203125, "ewc_loss_diag": 4.954636096954346e-07, "ewc_loss_parallel": 7.599592208862305e-07, "grad_norm": 4.604690074920654, "learning_rate": 6.273844849512505e-08, "loss": 0.7497, "mean_token_accuracy": 0.782461404800415, "num_tokens": 5723247.0, "step": 149 }, { "epoch": 0.019081541788576517, "ewc_loss": 0.00078582763671875, "ewc_loss_diag": 5.029141902923584e-07, "ewc_loss_parallel": 7.37607479095459e-07, "grad_norm": 4.441756248474121, "learning_rate": 6.316235693090292e-08, "loss": 0.7642, "mean_token_accuracy": 0.7787742018699646, "num_tokens": 5760305.0, "step": 150 }, { "epoch": 0.019208752067167028, "ewc_loss": 0.000774383544921875, "ewc_loss_diag": 5.029141902923584e-07, "ewc_loss_parallel": 7.264316082000732e-07, "grad_norm": 4.215742111206055, "learning_rate": 6.35862653666808e-08, "loss": 0.7645, "mean_token_accuracy": 0.7772437930107117, "num_tokens": 5800586.0, "step": 151 }, { "epoch": 0.01933596234575754, "ewc_loss": 0.000762939453125, "ewc_loss_diag": 5.103647708892822e-07, "ewc_loss_parallel": 7.115304470062256e-07, "grad_norm": 4.214138031005859, "learning_rate": 6.401017380245867e-08, "loss": 0.7562, "mean_token_accuracy": 0.7795295715332031, "num_tokens": 5840351.0, "step": 152 }, { "epoch": 0.019463172624348046, "ewc_loss": 0.00075531005859375, "ewc_loss_diag": 5.252659320831299e-07, "ewc_loss_parallel": 7.003545761108398e-07, "grad_norm": 4.265010833740234, "learning_rate": 6.443408223823654e-08, "loss": 0.7965, "mean_token_accuracy": 0.7706707119941711, "num_tokens": 5880162.0, "step": 153 }, { "epoch": 0.019590382902938557, "ewc_loss": 0.000728607177734375, "ewc_loss_diag": 5.364418029785156e-07, "ewc_loss_parallel": 6.742775440216064e-07, "grad_norm": 4.374281883239746, "learning_rate": 6.485799067401441e-08, "loss": 0.8526, "mean_token_accuracy": 0.7551860809326172, "num_tokens": 5922213.0, "step": 154 }, { "epoch": 0.019717593181529068, "ewc_loss": 0.0007171630859375, "ewc_loss_diag": 5.438923835754395e-07, "ewc_loss_parallel": 6.631016731262207e-07, "grad_norm": 4.267070293426514, "learning_rate": 6.528189910979228e-08, "loss": 0.7338, "mean_token_accuracy": 0.7854204177856445, "num_tokens": 5957461.0, "step": 155 }, { "epoch": 0.01984480346011958, "ewc_loss": 0.0006866455078125, "ewc_loss_diag": 5.513429641723633e-07, "ewc_loss_parallel": 6.332993507385254e-07, "grad_norm": 4.183377742767334, "learning_rate": 6.570580754557016e-08, "loss": 0.719, "mean_token_accuracy": 0.7857740521430969, "num_tokens": 5994675.0, "step": 156 }, { "epoch": 0.019972013738710086, "ewc_loss": 0.000682830810546875, "ewc_loss_diag": 5.587935447692871e-07, "ewc_loss_parallel": 6.258487701416016e-07, "grad_norm": 4.210832595825195, "learning_rate": 6.612971598134802e-08, "loss": 0.6823, "mean_token_accuracy": 0.801277756690979, "num_tokens": 6032514.0, "step": 157 }, { "epoch": 0.020099224017300597, "ewc_loss": 0.00067138671875, "ewc_loss_diag": 5.662441253662109e-07, "ewc_loss_parallel": 6.146728992462158e-07, "grad_norm": 4.2110419273376465, "learning_rate": 6.655362441712589e-08, "loss": 0.7544, "mean_token_accuracy": 0.7799386978149414, "num_tokens": 6069160.0, "step": 158 }, { "epoch": 0.020226434295891108, "ewc_loss": 0.000652313232421875, "ewc_loss_diag": 5.699694156646729e-07, "ewc_loss_parallel": 5.960464477539062e-07, "grad_norm": 4.265925407409668, "learning_rate": 6.697753285290376e-08, "loss": 0.8086, "mean_token_accuracy": 0.7644279599189758, "num_tokens": 6106652.0, "step": 159 }, { "epoch": 0.02035364457448162, "ewc_loss": 0.0006256103515625, "ewc_loss_diag": 5.811452865600586e-07, "ewc_loss_parallel": 5.699694156646729e-07, "grad_norm": 4.105129718780518, "learning_rate": 6.740144128868163e-08, "loss": 0.7427, "mean_token_accuracy": 0.7861591577529907, "num_tokens": 6143070.0, "step": 160 }, { "epoch": 0.020480854853072127, "ewc_loss": 0.00061798095703125, "ewc_loss_diag": 5.923211574554443e-07, "ewc_loss_parallel": 5.62518835067749e-07, "grad_norm": 3.913595199584961, "learning_rate": 6.78253497244595e-08, "loss": 0.7351, "mean_token_accuracy": 0.7865532636642456, "num_tokens": 6186495.0, "step": 161 }, { "epoch": 0.020608065131662637, "ewc_loss": 0.0006103515625, "ewc_loss_diag": 6.07222318649292e-07, "ewc_loss_parallel": 5.476176738739014e-07, "grad_norm": 4.13579797744751, "learning_rate": 6.824925816023738e-08, "loss": 0.7358, "mean_token_accuracy": 0.7857715487480164, "num_tokens": 6222454.0, "step": 162 }, { "epoch": 0.02073527541025315, "ewc_loss": 0.000606536865234375, "ewc_loss_diag": 6.221234798431396e-07, "ewc_loss_parallel": 5.438923835754395e-07, "grad_norm": 4.493397235870361, "learning_rate": 6.867316659601525e-08, "loss": 0.7481, "mean_token_accuracy": 0.7845291495323181, "num_tokens": 6253260.0, "step": 163 }, { "epoch": 0.02086248568884366, "ewc_loss": 0.0006103515625, "ewc_loss_diag": 6.370246410369873e-07, "ewc_loss_parallel": 5.438923835754395e-07, "grad_norm": 4.01668119430542, "learning_rate": 6.909707503179312e-08, "loss": 0.729, "mean_token_accuracy": 0.7899590134620667, "num_tokens": 6294773.0, "step": 164 }, { "epoch": 0.02098969596743417, "ewc_loss": 0.000591278076171875, "ewc_loss_diag": 6.51925802230835e-07, "ewc_loss_parallel": 5.252659320831299e-07, "grad_norm": 3.982788562774658, "learning_rate": 6.9520983467571e-08, "loss": 0.7726, "mean_token_accuracy": 0.7757415771484375, "num_tokens": 6335334.0, "step": 165 }, { "epoch": 0.021116906246024678, "ewc_loss": 0.000576019287109375, "ewc_loss_diag": 6.593763828277588e-07, "ewc_loss_parallel": 5.103647708892822e-07, "grad_norm": 3.9970552921295166, "learning_rate": 6.994489190334887e-08, "loss": 0.7081, "mean_token_accuracy": 0.7915183305740356, "num_tokens": 6372255.0, "step": 166 }, { "epoch": 0.02124411652461519, "ewc_loss": 0.0005645751953125, "ewc_loss_diag": 6.668269634246826e-07, "ewc_loss_parallel": 4.954636096954346e-07, "grad_norm": 3.836113452911377, "learning_rate": 7.036880033912674e-08, "loss": 0.6841, "mean_token_accuracy": 0.8004153966903687, "num_tokens": 6411971.0, "step": 167 }, { "epoch": 0.0213713268032057, "ewc_loss": 0.000553131103515625, "ewc_loss_diag": 6.817281246185303e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 4.1986799240112305, "learning_rate": 7.079270877490461e-08, "loss": 0.7546, "mean_token_accuracy": 0.7796882390975952, "num_tokens": 6450366.0, "step": 168 }, { "epoch": 0.02149853708179621, "ewc_loss": 0.000553131103515625, "ewc_loss_diag": 6.966292858123779e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 4.249285697937012, "learning_rate": 7.121661721068249e-08, "loss": 0.7174, "mean_token_accuracy": 0.7889552116394043, "num_tokens": 6482199.0, "step": 169 }, { "epoch": 0.021625747360386718, "ewc_loss": 0.00055694580078125, "ewc_loss_diag": 7.078051567077637e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 3.9425408840179443, "learning_rate": 7.164052564646036e-08, "loss": 0.699, "mean_token_accuracy": 0.7984463572502136, "num_tokens": 6519697.0, "step": 170 }, { "epoch": 0.02175295763897723, "ewc_loss": 0.00055694580078125, "ewc_loss_diag": 7.152557373046875e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 4.101984024047852, "learning_rate": 7.206443408223823e-08, "loss": 0.7591, "mean_token_accuracy": 0.778576135635376, "num_tokens": 6554687.0, "step": 171 }, { "epoch": 0.02188016791756774, "ewc_loss": 0.000553131103515625, "ewc_loss_diag": 7.301568984985352e-07, "ewc_loss_parallel": 4.805624485015869e-07, "grad_norm": 4.069310188293457, "learning_rate": 7.24883425180161e-08, "loss": 0.7221, "mean_token_accuracy": 0.782074511051178, "num_tokens": 6593066.0, "step": 172 }, { "epoch": 0.02200737819615825, "ewc_loss": 0.00054931640625, "ewc_loss_diag": 7.413327693939209e-07, "ewc_loss_parallel": 4.76837158203125e-07, "grad_norm": 3.7176527976989746, "learning_rate": 7.291225095379398e-08, "loss": 0.6799, "mean_token_accuracy": 0.8020479679107666, "num_tokens": 6635484.0, "step": 173 }, { "epoch": 0.02213458847474876, "ewc_loss": 0.00054931640625, "ewc_loss_diag": 7.525086402893066e-07, "ewc_loss_parallel": 4.731118679046631e-07, "grad_norm": 3.754798650741577, "learning_rate": 7.333615938957185e-08, "loss": 0.7727, "mean_token_accuracy": 0.7759313583374023, "num_tokens": 6677090.0, "step": 174 }, { "epoch": 0.02226179875333927, "ewc_loss": 0.00054931640625, "ewc_loss_diag": 7.7858567237854e-07, "ewc_loss_parallel": 4.7124922275543213e-07, "grad_norm": 4.1211323738098145, "learning_rate": 7.376006782534971e-08, "loss": 0.688, "mean_token_accuracy": 0.7964712977409363, "num_tokens": 6710250.0, "step": 175 }, { "epoch": 0.02238900903192978, "ewc_loss": 0.00055694580078125, "ewc_loss_diag": 7.897615432739258e-07, "ewc_loss_parallel": 4.76837158203125e-07, "grad_norm": 3.7459208965301514, "learning_rate": 7.418397626112758e-08, "loss": 0.7143, "mean_token_accuracy": 0.7887691259384155, "num_tokens": 6752998.0, "step": 176 }, { "epoch": 0.02251621931052029, "ewc_loss": 0.0005645751953125, "ewc_loss_diag": 8.009374141693115e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 3.8878612518310547, "learning_rate": 7.460788469690545e-08, "loss": 0.7153, "mean_token_accuracy": 0.7938848733901978, "num_tokens": 6789568.0, "step": 177 }, { "epoch": 0.0226434295891108, "ewc_loss": 0.000568389892578125, "ewc_loss_diag": 8.121132850646973e-07, "ewc_loss_parallel": 4.880130290985107e-07, "grad_norm": 4.368273735046387, "learning_rate": 7.503179313268333e-08, "loss": 0.756, "mean_token_accuracy": 0.7779221534729004, "num_tokens": 6822810.0, "step": 178 }, { "epoch": 0.02277063986770131, "ewc_loss": 0.000583648681640625, "ewc_loss_diag": 8.23289155960083e-07, "ewc_loss_parallel": 4.991888999938965e-07, "grad_norm": 3.869462251663208, "learning_rate": 7.54557015684612e-08, "loss": 0.6951, "mean_token_accuracy": 0.7939528226852417, "num_tokens": 6861598.0, "step": 179 }, { "epoch": 0.02289785014629182, "ewc_loss": 0.000583648681640625, "ewc_loss_diag": 8.493661880493164e-07, "ewc_loss_parallel": 4.991888999938965e-07, "grad_norm": 3.847465753555298, "learning_rate": 7.587961000423907e-08, "loss": 0.7172, "mean_token_accuracy": 0.7909437417984009, "num_tokens": 6900220.0, "step": 180 }, { "epoch": 0.02302506042488233, "ewc_loss": 0.0005950927734375, "ewc_loss_diag": 8.642673492431641e-07, "ewc_loss_parallel": 5.066394805908203e-07, "grad_norm": 3.8676936626434326, "learning_rate": 7.630351844001694e-08, "loss": 0.8257, "mean_token_accuracy": 0.7600595951080322, "num_tokens": 6946983.0, "step": 181 }, { "epoch": 0.02315227070347284, "ewc_loss": 0.00060272216796875, "ewc_loss_diag": 8.828938007354736e-07, "ewc_loss_parallel": 5.140900611877441e-07, "grad_norm": 4.135307788848877, "learning_rate": 7.672742687579482e-08, "loss": 0.7306, "mean_token_accuracy": 0.7872251272201538, "num_tokens": 6986019.0, "step": 182 }, { "epoch": 0.02327948098206335, "ewc_loss": 0.00061798095703125, "ewc_loss_diag": 9.5367431640625e-07, "ewc_loss_parallel": 5.21540641784668e-07, "grad_norm": 4.132412433624268, "learning_rate": 7.715133531157269e-08, "loss": 0.7041, "mean_token_accuracy": 0.7889752984046936, "num_tokens": 7020722.0, "step": 183 }, { "epoch": 0.02340669126065386, "ewc_loss": 0.000644683837890625, "ewc_loss_diag": 9.611248970031738e-07, "ewc_loss_parallel": 5.476176738739014e-07, "grad_norm": 3.9530041217803955, "learning_rate": 7.757524374735056e-08, "loss": 0.7752, "mean_token_accuracy": 0.7748453617095947, "num_tokens": 7062691.0, "step": 184 }, { "epoch": 0.02353390153924437, "ewc_loss": 0.00066375732421875, "ewc_loss_diag": 9.685754776000977e-07, "ewc_loss_parallel": 5.699694156646729e-07, "grad_norm": 3.9998230934143066, "learning_rate": 7.799915218312844e-08, "loss": 0.7231, "mean_token_accuracy": 0.7890151739120483, "num_tokens": 7101090.0, "step": 185 }, { "epoch": 0.02366111181783488, "ewc_loss": 0.0006866455078125, "ewc_loss_diag": 9.760260581970215e-07, "ewc_loss_parallel": 5.885958671569824e-07, "grad_norm": 3.926255702972412, "learning_rate": 7.842306061890631e-08, "loss": 0.6964, "mean_token_accuracy": 0.7915184497833252, "num_tokens": 7139891.0, "step": 186 }, { "epoch": 0.023788322096425393, "ewc_loss": 0.000698089599609375, "ewc_loss_diag": 9.909272193908691e-07, "ewc_loss_parallel": 5.997717380523682e-07, "grad_norm": 3.9824471473693848, "learning_rate": 7.88469690546842e-08, "loss": 0.7444, "mean_token_accuracy": 0.7817773818969727, "num_tokens": 7179501.0, "step": 187 }, { "epoch": 0.0239155323750159, "ewc_loss": 0.000720977783203125, "ewc_loss_diag": 9.5367431640625e-07, "ewc_loss_parallel": 6.258487701416016e-07, "grad_norm": 4.068840980529785, "learning_rate": 7.927087749046207e-08, "loss": 0.7137, "mean_token_accuracy": 0.7888346910476685, "num_tokens": 7213809.0, "step": 188 }, { "epoch": 0.02404274265360641, "ewc_loss": 0.000736236572265625, "ewc_loss_diag": 9.611248970031738e-07, "ewc_loss_parallel": 6.407499313354492e-07, "grad_norm": 3.970604181289673, "learning_rate": 7.969478592623994e-08, "loss": 0.7886, "mean_token_accuracy": 0.7671737670898438, "num_tokens": 7254493.0, "step": 189 }, { "epoch": 0.024169952932196922, "ewc_loss": 0.000762939453125, "ewc_loss_diag": 9.685754776000977e-07, "ewc_loss_parallel": 6.668269634246826e-07, "grad_norm": 4.162939548492432, "learning_rate": 8.011869436201781e-08, "loss": 0.6462, "mean_token_accuracy": 0.8091422915458679, "num_tokens": 7292643.0, "step": 190 }, { "epoch": 0.024297163210787433, "ewc_loss": 0.000804901123046875, "ewc_loss_diag": 9.760260581970215e-07, "ewc_loss_parallel": 7.040798664093018e-07, "grad_norm": 4.1396098136901855, "learning_rate": 8.054260279779568e-08, "loss": 0.7691, "mean_token_accuracy": 0.7763917446136475, "num_tokens": 7329810.0, "step": 191 }, { "epoch": 0.02442437348937794, "ewc_loss": 0.000820159912109375, "ewc_loss_diag": 9.834766387939453e-07, "ewc_loss_parallel": 7.227063179016113e-07, "grad_norm": 3.9495084285736084, "learning_rate": 8.096651123357356e-08, "loss": 0.7551, "mean_token_accuracy": 0.7776923179626465, "num_tokens": 7367630.0, "step": 192 }, { "epoch": 0.02455158376796845, "ewc_loss": 0.000843048095703125, "ewc_loss_diag": 9.98377799987793e-07, "ewc_loss_parallel": 7.450580596923828e-07, "grad_norm": 4.230428218841553, "learning_rate": 8.139041966935143e-08, "loss": 0.7364, "mean_token_accuracy": 0.7830501794815063, "num_tokens": 7411580.0, "step": 193 }, { "epoch": 0.024678794046558962, "ewc_loss": 0.000885009765625, "ewc_loss_diag": 1.0058283805847168e-06, "ewc_loss_parallel": 7.860362529754639e-07, "grad_norm": 4.215463161468506, "learning_rate": 8.181432810512929e-08, "loss": 0.6908, "mean_token_accuracy": 0.7956995964050293, "num_tokens": 7451631.0, "step": 194 }, { "epoch": 0.024806004325149473, "ewc_loss": 0.00091552734375, "ewc_loss_diag": 1.0132789611816406e-06, "ewc_loss_parallel": 8.158385753631592e-07, "grad_norm": 3.9079020023345947, "learning_rate": 8.223823654090716e-08, "loss": 0.6995, "mean_token_accuracy": 0.7921697497367859, "num_tokens": 7493645.0, "step": 195 }, { "epoch": 0.024933214603739984, "ewc_loss": 0.00095367431640625, "ewc_loss_diag": 1.0281801223754883e-06, "ewc_loss_parallel": 8.493661880493164e-07, "grad_norm": 4.450375080108643, "learning_rate": 8.266214497668503e-08, "loss": 0.66, "mean_token_accuracy": 0.8063863515853882, "num_tokens": 7538042.0, "step": 196 }, { "epoch": 0.02506042488233049, "ewc_loss": 0.0009765625, "ewc_loss_diag": 1.0505318641662598e-06, "ewc_loss_parallel": 8.754432201385498e-07, "grad_norm": 4.016750335693359, "learning_rate": 8.30860534124629e-08, "loss": 0.7222, "mean_token_accuracy": 0.7860852479934692, "num_tokens": 7575374.0, "step": 197 }, { "epoch": 0.025187635160921002, "ewc_loss": 0.001007080078125, "ewc_loss_diag": 1.0579824447631836e-06, "ewc_loss_parallel": 9.015202522277832e-07, "grad_norm": 4.283174991607666, "learning_rate": 8.350996184824078e-08, "loss": 0.7148, "mean_token_accuracy": 0.7870862483978271, "num_tokens": 7606880.0, "step": 198 }, { "epoch": 0.025314845439511513, "ewc_loss": 0.00104522705078125, "ewc_loss_diag": 1.0728836059570312e-06, "ewc_loss_parallel": 9.387731552124023e-07, "grad_norm": 4.554825305938721, "learning_rate": 8.393387028401865e-08, "loss": 0.6925, "mean_token_accuracy": 0.7966973781585693, "num_tokens": 7644840.0, "step": 199 }, { "epoch": 0.025442055718102024, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.0952353477478027e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 4.119286060333252, "learning_rate": 8.435777871979652e-08, "loss": 0.733, "mean_token_accuracy": 0.7825649976730347, "num_tokens": 7683856.0, "step": 200 }, { "epoch": 0.02556926599669253, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 1.1026859283447266e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 5.37406587600708, "learning_rate": 8.47816871555744e-08, "loss": 0.6912, "mean_token_accuracy": 0.7958367466926575, "num_tokens": 7715306.0, "step": 201 }, { "epoch": 0.025696476275283042, "ewc_loss": 0.00115203857421875, "ewc_loss_diag": 1.1399388313293457e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 4.310070514678955, "learning_rate": 8.520559559135227e-08, "loss": 0.6512, "mean_token_accuracy": 0.8057568073272705, "num_tokens": 7752442.0, "step": 202 }, { "epoch": 0.025823686553873553, "ewc_loss": 0.0011749267578125, "ewc_loss_diag": 1.259148120880127e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 4.858307361602783, "learning_rate": 8.562950402713014e-08, "loss": 0.7353, "mean_token_accuracy": 0.7848973274230957, "num_tokens": 7799891.0, "step": 203 }, { "epoch": 0.025950896832464064, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.2665987014770508e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 4.389342784881592, "learning_rate": 8.605341246290801e-08, "loss": 0.7235, "mean_token_accuracy": 0.7835657596588135, "num_tokens": 7830696.0, "step": 204 }, { "epoch": 0.026078107111054575, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.2814998626708984e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 4.038605690002441, "learning_rate": 8.647732089868589e-08, "loss": 0.6945, "mean_token_accuracy": 0.7960413098335266, "num_tokens": 7870968.0, "step": 205 }, { "epoch": 0.026205317389645082, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 1.2442469596862793e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 4.135344505310059, "learning_rate": 8.690122933446376e-08, "loss": 0.7184, "mean_token_accuracy": 0.7887403964996338, "num_tokens": 7914418.0, "step": 206 }, { "epoch": 0.026332527668235593, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 1.2516975402832031e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 3.9742345809936523, "learning_rate": 8.732513777024163e-08, "loss": 0.6556, "mean_token_accuracy": 0.810113251209259, "num_tokens": 7954544.0, "step": 207 }, { "epoch": 0.026459737946826104, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 1.2665987014770508e-06, "ewc_loss_parallel": 1.0728836059570312e-06, "grad_norm": 4.599030017852783, "learning_rate": 8.77490462060195e-08, "loss": 0.7048, "mean_token_accuracy": 0.7946653366088867, "num_tokens": 7997373.0, "step": 208 }, { "epoch": 0.026586948225416615, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.2814998626708984e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 4.602856159210205, "learning_rate": 8.817295464179738e-08, "loss": 0.6681, "mean_token_accuracy": 0.7971355319023132, "num_tokens": 8033314.0, "step": 209 }, { "epoch": 0.026714158504007122, "ewc_loss": 0.001251220703125, "ewc_loss_diag": 1.296401023864746e-06, "ewc_loss_parallel": 1.125037670135498e-06, "grad_norm": 5.079585552215576, "learning_rate": 8.859686307757525e-08, "loss": 0.687, "mean_token_accuracy": 0.7964198589324951, "num_tokens": 8069362.0, "step": 210 }, { "epoch": 0.026841368782597633, "ewc_loss": 0.00127410888671875, "ewc_loss_diag": 1.3113021850585938e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 4.5916571617126465, "learning_rate": 8.902077151335312e-08, "loss": 0.7199, "mean_token_accuracy": 0.7849040031433105, "num_tokens": 8109126.0, "step": 211 }, { "epoch": 0.026968579061188144, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 1.341104507446289e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 4.4697346687316895, "learning_rate": 8.944467994913098e-08, "loss": 0.7733, "mean_token_accuracy": 0.776383638381958, "num_tokens": 8143083.0, "step": 212 }, { "epoch": 0.027095789339778655, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 1.3634562492370605e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 4.690909385681152, "learning_rate": 8.986858838490885e-08, "loss": 0.6231, "mean_token_accuracy": 0.8089545965194702, "num_tokens": 8176636.0, "step": 213 }, { "epoch": 0.027222999618369163, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 1.3634562492370605e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 4.473193168640137, "learning_rate": 9.029249682068673e-08, "loss": 0.6828, "mean_token_accuracy": 0.7958294749259949, "num_tokens": 8220262.0, "step": 214 }, { "epoch": 0.027350209896959674, "ewc_loss": 0.00127410888671875, "ewc_loss_diag": 1.3783574104309082e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 5.377935886383057, "learning_rate": 9.07164052564646e-08, "loss": 0.7011, "mean_token_accuracy": 0.7936660051345825, "num_tokens": 8256303.0, "step": 215 }, { "epoch": 0.027477420175550184, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 1.385807991027832e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.132225036621094, "learning_rate": 9.114031369224247e-08, "loss": 0.6669, "mean_token_accuracy": 0.8036394119262695, "num_tokens": 8293888.0, "step": 216 }, { "epoch": 0.027604630454140695, "ewc_loss": 0.001312255859375, "ewc_loss_diag": 1.4081597328186035e-06, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 4.3190741539001465, "learning_rate": 9.156422212802034e-08, "loss": 0.6282, "mean_token_accuracy": 0.8109481334686279, "num_tokens": 8335368.0, "step": 217 }, { "epoch": 0.027731840732731206, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 1.4156103134155273e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 5.266815185546875, "learning_rate": 9.198813056379822e-08, "loss": 0.6826, "mean_token_accuracy": 0.79936283826828, "num_tokens": 8370338.0, "step": 218 }, { "epoch": 0.027859051011321714, "ewc_loss": 0.00127410888671875, "ewc_loss_diag": 1.430511474609375e-06, "ewc_loss_parallel": 1.1324882507324219e-06, "grad_norm": 4.403353691101074, "learning_rate": 9.241203899957609e-08, "loss": 0.6677, "mean_token_accuracy": 0.7994523048400879, "num_tokens": 8409904.0, "step": 219 }, { "epoch": 0.027986261289912225, "ewc_loss": 0.0012664794921875, "ewc_loss_diag": 1.4454126358032227e-06, "ewc_loss_parallel": 1.125037670135498e-06, "grad_norm": 4.431788921356201, "learning_rate": 9.283594743535396e-08, "loss": 0.7951, "mean_token_accuracy": 0.7675654888153076, "num_tokens": 8446579.0, "step": 220 }, { "epoch": 0.028113471568502735, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 1.4528632164001465e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 4.088932991027832, "learning_rate": 9.325985587113183e-08, "loss": 0.6425, "mean_token_accuracy": 0.8096162676811218, "num_tokens": 8483663.0, "step": 221 }, { "epoch": 0.028240681847093246, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.4677643775939941e-06, "ewc_loss_parallel": 1.0728836059570312e-06, "grad_norm": 5.411437511444092, "learning_rate": 9.368376430690971e-08, "loss": 0.6727, "mean_token_accuracy": 0.7987207174301147, "num_tokens": 8516346.0, "step": 222 }, { "epoch": 0.028367892125683754, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 1.5050172805786133e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 4.234001159667969, "learning_rate": 9.410767274268758e-08, "loss": 0.6687, "mean_token_accuracy": 0.8022159337997437, "num_tokens": 8556914.0, "step": 223 }, { "epoch": 0.028495102404274265, "ewc_loss": 0.00124359130859375, "ewc_loss_diag": 1.5348196029663086e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 4.312453746795654, "learning_rate": 9.453158117846545e-08, "loss": 0.6574, "mean_token_accuracy": 0.8011772632598877, "num_tokens": 8596548.0, "step": 224 }, { "epoch": 0.028622312682864776, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.5050172805786133e-06, "ewc_loss_parallel": 1.0654330253601074e-06, "grad_norm": 5.244631290435791, "learning_rate": 9.495548961424333e-08, "loss": 0.6879, "mean_token_accuracy": 0.7988945245742798, "num_tokens": 8634366.0, "step": 225 }, { "epoch": 0.028749522961455286, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.5124678611755371e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 4.577944278717041, "learning_rate": 9.53793980500212e-08, "loss": 0.6275, "mean_token_accuracy": 0.8101421594619751, "num_tokens": 8666480.0, "step": 226 }, { "epoch": 0.028876733240045797, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.5869736671447754e-06, "ewc_loss_parallel": 1.0728836059570312e-06, "grad_norm": 4.708883762359619, "learning_rate": 9.580330648579907e-08, "loss": 0.7772, "mean_token_accuracy": 0.7750471234321594, "num_tokens": 8705880.0, "step": 227 }, { "epoch": 0.029003943518636305, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 1.5944242477416992e-06, "ewc_loss_parallel": 1.0654330253601074e-06, "grad_norm": 4.706445217132568, "learning_rate": 9.622721492157694e-08, "loss": 0.6389, "mean_token_accuracy": 0.8083744049072266, "num_tokens": 8743772.0, "step": 228 }, { "epoch": 0.029131153797226816, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 1.6093254089355469e-06, "ewc_loss_parallel": 1.0654330253601074e-06, "grad_norm": 4.728447437286377, "learning_rate": 9.665112335735482e-08, "loss": 0.6638, "mean_token_accuracy": 0.7972354292869568, "num_tokens": 8775919.0, "step": 229 }, { "epoch": 0.029258364075817327, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 1.5720725059509277e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 5.579430103302002, "learning_rate": 9.707503179313267e-08, "loss": 0.6122, "mean_token_accuracy": 0.8142833709716797, "num_tokens": 8809336.0, "step": 230 }, { "epoch": 0.029385574354407838, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.601874828338623e-06, "ewc_loss_parallel": 1.0728836059570312e-06, "grad_norm": 5.004061698913574, "learning_rate": 9.749894022891055e-08, "loss": 0.6654, "mean_token_accuracy": 0.8002310991287231, "num_tokens": 8845952.0, "step": 231 }, { "epoch": 0.029512784632998345, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 1.6242265701293945e-06, "ewc_loss_parallel": 1.0654330253601074e-06, "grad_norm": 4.802040100097656, "learning_rate": 9.792284866468842e-08, "loss": 0.6854, "mean_token_accuracy": 0.7920489311218262, "num_tokens": 8889801.0, "step": 232 }, { "epoch": 0.029639994911588856, "ewc_loss": 0.00119781494140625, "ewc_loss_diag": 1.6316771507263184e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 4.314411163330078, "learning_rate": 9.834675710046629e-08, "loss": 0.6442, "mean_token_accuracy": 0.8033344745635986, "num_tokens": 8925429.0, "step": 233 }, { "epoch": 0.029767205190179367, "ewc_loss": 0.0011749267578125, "ewc_loss_diag": 1.6316771507263184e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 5.088131427764893, "learning_rate": 9.877066553624416e-08, "loss": 0.6773, "mean_token_accuracy": 0.7957332730293274, "num_tokens": 8963360.0, "step": 234 }, { "epoch": 0.029894415468769878, "ewc_loss": 0.0011749267578125, "ewc_loss_diag": 1.6391277313232422e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 5.946727752685547, "learning_rate": 9.919457397202204e-08, "loss": 0.6881, "mean_token_accuracy": 0.7974632382392883, "num_tokens": 8998314.0, "step": 235 }, { "epoch": 0.030021625747360385, "ewc_loss": 0.001190185546875, "ewc_loss_diag": 1.6540288925170898e-06, "ewc_loss_parallel": 1.0207295417785645e-06, "grad_norm": 4.792496681213379, "learning_rate": 9.961848240779991e-08, "loss": 0.7579, "mean_token_accuracy": 0.7768839001655579, "num_tokens": 9034804.0, "step": 236 }, { "epoch": 0.030148836025950896, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 1.6614794731140137e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 4.725142955780029, "learning_rate": 1.0004239084357778e-07, "loss": 0.6805, "mean_token_accuracy": 0.7970014810562134, "num_tokens": 9070840.0, "step": 237 }, { "epoch": 0.030276046304541407, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 1.691281795501709e-06, "ewc_loss_parallel": 9.98377799987793e-07, "grad_norm": 4.707487106323242, "learning_rate": 1.0046629927935566e-07, "loss": 0.6621, "mean_token_accuracy": 0.803016185760498, "num_tokens": 9109370.0, "step": 238 }, { "epoch": 0.030403256583131918, "ewc_loss": 0.00115203857421875, "ewc_loss_diag": 1.6987323760986328e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.838991165161133, "learning_rate": 1.0089020771513353e-07, "loss": 0.6057, "mean_token_accuracy": 0.8185278177261353, "num_tokens": 9148191.0, "step": 239 }, { "epoch": 0.03053046686172243, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 1.6838312149047852e-06, "ewc_loss_parallel": 9.98377799987793e-07, "grad_norm": 4.932633876800537, "learning_rate": 1.013141161509114e-07, "loss": 0.6728, "mean_token_accuracy": 0.7995984554290771, "num_tokens": 9182923.0, "step": 240 }, { "epoch": 0.030657677140312936, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 1.6987323760986328e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 4.825782775878906, "learning_rate": 1.0173802458668927e-07, "loss": 0.6717, "mean_token_accuracy": 0.7962172031402588, "num_tokens": 9215882.0, "step": 241 }, { "epoch": 0.030784887418903447, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 1.7136335372924805e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 6.405088424682617, "learning_rate": 1.0216193302246715e-07, "loss": 0.7042, "mean_token_accuracy": 0.7848544716835022, "num_tokens": 9249718.0, "step": 242 }, { "epoch": 0.030912097697493958, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 1.7136335372924805e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 4.430290222167969, "learning_rate": 1.0258584145824502e-07, "loss": 0.6261, "mean_token_accuracy": 0.8117128610610962, "num_tokens": 9291146.0, "step": 243 }, { "epoch": 0.03103930797608447, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 1.7210841178894043e-06, "ewc_loss_parallel": 9.611248970031738e-07, "grad_norm": 5.247747898101807, "learning_rate": 1.0300974989402289e-07, "loss": 0.6178, "mean_token_accuracy": 0.81182861328125, "num_tokens": 9326403.0, "step": 244 }, { "epoch": 0.031166518254674976, "ewc_loss": 0.00112152099609375, "ewc_loss_diag": 1.735985279083252e-06, "ewc_loss_parallel": 9.499490261077881e-07, "grad_norm": 4.882724761962891, "learning_rate": 1.0343365832980076e-07, "loss": 0.671, "mean_token_accuracy": 0.7992497086524963, "num_tokens": 9368491.0, "step": 245 }, { "epoch": 0.03129372853326549, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 1.735985279083252e-06, "ewc_loss_parallel": 9.313225746154785e-07, "grad_norm": 5.117701053619385, "learning_rate": 1.0385756676557864e-07, "loss": 0.6667, "mean_token_accuracy": 0.8013550043106079, "num_tokens": 9409041.0, "step": 246 }, { "epoch": 0.031420938811855995, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 1.7434358596801758e-06, "ewc_loss_parallel": 9.313225746154785e-07, "grad_norm": 4.953301906585693, "learning_rate": 1.0428147520135651e-07, "loss": 0.6397, "mean_token_accuracy": 0.8078907132148743, "num_tokens": 9448591.0, "step": 247 }, { "epoch": 0.03154814909044651, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.7434358596801758e-06, "ewc_loss_parallel": 9.238719940185547e-07, "grad_norm": 4.679422855377197, "learning_rate": 1.0470538363713437e-07, "loss": 0.6698, "mean_token_accuracy": 0.7973447442054749, "num_tokens": 9486604.0, "step": 248 }, { "epoch": 0.031675359369037016, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 1.7583370208740234e-06, "ewc_loss_parallel": 9.164214134216309e-07, "grad_norm": 5.643101692199707, "learning_rate": 1.0512929207291224e-07, "loss": 0.6758, "mean_token_accuracy": 0.7928627133369446, "num_tokens": 9525285.0, "step": 249 }, { "epoch": 0.03180256964762753, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.7583370208740234e-06, "ewc_loss_parallel": 9.201467037200928e-07, "grad_norm": 6.139242172241211, "learning_rate": 1.0555320050869011e-07, "loss": 0.656, "mean_token_accuracy": 0.7979235649108887, "num_tokens": 9557898.0, "step": 250 }, { "epoch": 0.03192977992621804, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 1.7583370208740234e-06, "ewc_loss_parallel": 9.5367431640625e-07, "grad_norm": 5.540950775146484, "learning_rate": 1.0597710894446799e-07, "loss": 0.7156, "mean_token_accuracy": 0.782342255115509, "num_tokens": 9590438.0, "step": 251 }, { "epoch": 0.032056990204808546, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 1.8253922462463379e-06, "ewc_loss_parallel": 9.462237358093262e-07, "grad_norm": 6.086142063140869, "learning_rate": 1.0640101738024586e-07, "loss": 0.626, "mean_token_accuracy": 0.8108742833137512, "num_tokens": 9631402.0, "step": 252 }, { "epoch": 0.03218420048339906, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 1.780688762664795e-06, "ewc_loss_parallel": 9.387731552124023e-07, "grad_norm": 5.0522637367248535, "learning_rate": 1.0682492581602373e-07, "loss": 0.6073, "mean_token_accuracy": 0.817456841468811, "num_tokens": 9663702.0, "step": 253 }, { "epoch": 0.03231141076198957, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.8402934074401855e-06, "ewc_loss_parallel": 9.015202522277832e-07, "grad_norm": 5.3593854904174805, "learning_rate": 1.072488342518016e-07, "loss": 0.5951, "mean_token_accuracy": 0.815717339515686, "num_tokens": 9698969.0, "step": 254 }, { "epoch": 0.03243862104058008, "ewc_loss": 0.00104522705078125, "ewc_loss_diag": 1.7955899238586426e-06, "ewc_loss_parallel": 8.642673492431641e-07, "grad_norm": 4.622766494750977, "learning_rate": 1.0767274268757948e-07, "loss": 0.6647, "mean_token_accuracy": 0.7981338500976562, "num_tokens": 9738075.0, "step": 255 }, { "epoch": 0.03256583131917059, "ewc_loss": 0.0010223388671875, "ewc_loss_diag": 1.7955899238586426e-06, "ewc_loss_parallel": 8.419156074523926e-07, "grad_norm": 4.5129714012146, "learning_rate": 1.0809665112335735e-07, "loss": 0.6695, "mean_token_accuracy": 0.8001359701156616, "num_tokens": 9782675.0, "step": 256 }, { "epoch": 0.0326930415977611, "ewc_loss": 0.00101470947265625, "ewc_loss_diag": 1.8253922462463379e-06, "ewc_loss_parallel": 8.344650268554688e-07, "grad_norm": 5.976316928863525, "learning_rate": 1.0852055955913522e-07, "loss": 0.57, "mean_token_accuracy": 0.8254082202911377, "num_tokens": 9820585.0, "step": 257 }, { "epoch": 0.03282025187635161, "ewc_loss": 0.00104522705078125, "ewc_loss_diag": 1.8402934074401855e-06, "ewc_loss_parallel": 8.568167686462402e-07, "grad_norm": 4.744744777679443, "learning_rate": 1.089444679949131e-07, "loss": 0.6943, "mean_token_accuracy": 0.7917486429214478, "num_tokens": 9860693.0, "step": 258 }, { "epoch": 0.03294746215494212, "ewc_loss": 0.00103759765625, "ewc_loss_diag": 1.8328428268432617e-06, "ewc_loss_parallel": 8.493661880493164e-07, "grad_norm": 5.192368030548096, "learning_rate": 1.0936837643069097e-07, "loss": 0.6276, "mean_token_accuracy": 0.810333788394928, "num_tokens": 9902629.0, "step": 259 }, { "epoch": 0.03307467243353263, "ewc_loss": 0.00103759765625, "ewc_loss_diag": 1.8402934074401855e-06, "ewc_loss_parallel": 8.493661880493164e-07, "grad_norm": 4.956759929656982, "learning_rate": 1.0979228486646884e-07, "loss": 0.5997, "mean_token_accuracy": 0.8144351243972778, "num_tokens": 9937304.0, "step": 260 }, { "epoch": 0.03320188271212314, "ewc_loss": 0.00103759765625, "ewc_loss_diag": 1.8477439880371094e-06, "ewc_loss_parallel": 8.530914783477783e-07, "grad_norm": 5.82897424697876, "learning_rate": 1.1021619330224671e-07, "loss": 0.7326, "mean_token_accuracy": 0.7802319526672363, "num_tokens": 9976911.0, "step": 261 }, { "epoch": 0.03332909299071365, "ewc_loss": 0.00106048583984375, "ewc_loss_diag": 1.862645149230957e-06, "ewc_loss_parallel": 8.754432201385498e-07, "grad_norm": 5.128118991851807, "learning_rate": 1.1064010173802458e-07, "loss": 0.6673, "mean_token_accuracy": 0.7994992136955261, "num_tokens": 10015740.0, "step": 262 }, { "epoch": 0.03345630326930416, "ewc_loss": 0.0010528564453125, "ewc_loss_diag": 1.8849968910217285e-06, "ewc_loss_parallel": 8.605420589447021e-07, "grad_norm": 5.273438930511475, "learning_rate": 1.1106401017380246e-07, "loss": 0.6045, "mean_token_accuracy": 0.8166487216949463, "num_tokens": 10053054.0, "step": 263 }, { "epoch": 0.03358351354789467, "ewc_loss": 0.00104522705078125, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 8.568167686462402e-07, "grad_norm": 7.353725910186768, "learning_rate": 1.1148791860958033e-07, "loss": 0.7005, "mean_token_accuracy": 0.7862496376037598, "num_tokens": 10082088.0, "step": 264 }, { "epoch": 0.03371072382648518, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 8.940696716308594e-07, "grad_norm": 5.206089973449707, "learning_rate": 1.119118270453582e-07, "loss": 0.6738, "mean_token_accuracy": 0.7959926128387451, "num_tokens": 10121122.0, "step": 265 }, { "epoch": 0.03383793410507569, "ewc_loss": 0.0010528564453125, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 8.642673492431641e-07, "grad_norm": 6.524083614349365, "learning_rate": 1.1233573548113607e-07, "loss": 0.6681, "mean_token_accuracy": 0.7992916703224182, "num_tokens": 10161017.0, "step": 266 }, { "epoch": 0.0339651443836662, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.717179298400879e-07, "grad_norm": 4.3290486335754395, "learning_rate": 1.1275964391691393e-07, "loss": 0.6343, "mean_token_accuracy": 0.8093597292900085, "num_tokens": 10204913.0, "step": 267 }, { "epoch": 0.03409235466225671, "ewc_loss": 0.001007080078125, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.158385753631592e-07, "grad_norm": 5.414713382720947, "learning_rate": 1.131835523526918e-07, "loss": 0.6821, "mean_token_accuracy": 0.7916994094848633, "num_tokens": 10245154.0, "step": 268 }, { "epoch": 0.03421956494084722, "ewc_loss": 0.001007080078125, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.158385753631592e-07, "grad_norm": 6.800817012786865, "learning_rate": 1.1360746078846968e-07, "loss": 0.6893, "mean_token_accuracy": 0.7915288805961609, "num_tokens": 10283006.0, "step": 269 }, { "epoch": 0.03434677521943773, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.754432201385498e-07, "grad_norm": 5.957331657409668, "learning_rate": 1.1403136922424755e-07, "loss": 0.6488, "mean_token_accuracy": 0.8023949265480042, "num_tokens": 10316689.0, "step": 270 }, { "epoch": 0.03447398549802824, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.866190910339355e-07, "grad_norm": 5.9050469398498535, "learning_rate": 1.1445527766002542e-07, "loss": 0.6435, "mean_token_accuracy": 0.8022632598876953, "num_tokens": 10358145.0, "step": 271 }, { "epoch": 0.03460119577661875, "ewc_loss": 0.0010528564453125, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.605420589447021e-07, "grad_norm": 5.387948036193848, "learning_rate": 1.148791860958033e-07, "loss": 0.744, "mean_token_accuracy": 0.7777427434921265, "num_tokens": 10393424.0, "step": 272 }, { "epoch": 0.034728406055209264, "ewc_loss": 0.00103759765625, "ewc_loss_diag": 1.952052116394043e-06, "ewc_loss_parallel": 8.381903171539307e-07, "grad_norm": 4.77749490737915, "learning_rate": 1.1530309453158117e-07, "loss": 0.6586, "mean_token_accuracy": 0.8004553914070129, "num_tokens": 10433560.0, "step": 273 }, { "epoch": 0.03485561633379977, "ewc_loss": 0.00099945068359375, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.083879947662354e-07, "grad_norm": 5.131369590759277, "learning_rate": 1.1572700296735904e-07, "loss": 0.6554, "mean_token_accuracy": 0.8005056381225586, "num_tokens": 10472024.0, "step": 274 }, { "epoch": 0.03498282661239028, "ewc_loss": 0.001007080078125, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.158385753631592e-07, "grad_norm": 5.551281929016113, "learning_rate": 1.1615091140313691e-07, "loss": 0.6939, "mean_token_accuracy": 0.7900619506835938, "num_tokens": 10510874.0, "step": 275 }, { "epoch": 0.03511003689098079, "ewc_loss": 0.00102996826171875, "ewc_loss_diag": 1.952052116394043e-06, "ewc_loss_parallel": 8.344650268554688e-07, "grad_norm": 5.120694637298584, "learning_rate": 1.1657481983891479e-07, "loss": 0.6317, "mean_token_accuracy": 0.8085429072380066, "num_tokens": 10552411.0, "step": 276 }, { "epoch": 0.0352372471695713, "ewc_loss": 0.00103759765625, "ewc_loss_diag": 1.9669532775878906e-06, "ewc_loss_parallel": 8.381903171539307e-07, "grad_norm": 4.823660850524902, "learning_rate": 1.1699872827469266e-07, "loss": 0.6474, "mean_token_accuracy": 0.8048595786094666, "num_tokens": 10591537.0, "step": 277 }, { "epoch": 0.03536445744816181, "ewc_loss": 0.0010223388671875, "ewc_loss_diag": 1.9818544387817383e-06, "ewc_loss_parallel": 8.23289155960083e-07, "grad_norm": 5.820143222808838, "learning_rate": 1.1742263671047053e-07, "loss": 0.6562, "mean_token_accuracy": 0.8021621108055115, "num_tokens": 10632228.0, "step": 278 }, { "epoch": 0.03549166772675232, "ewc_loss": 0.0010528564453125, "ewc_loss_diag": 1.9818544387817383e-06, "ewc_loss_parallel": 8.493661880493164e-07, "grad_norm": 6.85934591293335, "learning_rate": 1.178465451462484e-07, "loss": 0.7115, "mean_token_accuracy": 0.7832387685775757, "num_tokens": 10672829.0, "step": 279 }, { "epoch": 0.03561887800534283, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 8.977949619293213e-07, "grad_norm": 5.735762596130371, "learning_rate": 1.1827045358202628e-07, "loss": 0.6855, "mean_token_accuracy": 0.7983343601226807, "num_tokens": 10706146.0, "step": 280 }, { "epoch": 0.035746088283933344, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 8.903443813323975e-07, "grad_norm": 5.967990875244141, "learning_rate": 1.1869436201780415e-07, "loss": 0.6025, "mean_token_accuracy": 0.8182865381240845, "num_tokens": 10744112.0, "step": 281 }, { "epoch": 0.03587329856252385, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 2.0116567611694336e-06, "ewc_loss_parallel": 8.791685104370117e-07, "grad_norm": 5.587921142578125, "learning_rate": 1.1911827045358202e-07, "loss": 0.633, "mean_token_accuracy": 0.8127753734588623, "num_tokens": 10783726.0, "step": 282 }, { "epoch": 0.03600050884111436, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 8.605420589447021e-07, "grad_norm": 6.618085861206055, "learning_rate": 1.195421788893599e-07, "loss": 0.6976, "mean_token_accuracy": 0.7922855019569397, "num_tokens": 10824476.0, "step": 283 }, { "epoch": 0.036127719119704874, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 8.866190910339355e-07, "grad_norm": 7.315528869628906, "learning_rate": 1.1996608732513778e-07, "loss": 0.6573, "mean_token_accuracy": 0.8013728260993958, "num_tokens": 10859997.0, "step": 284 }, { "epoch": 0.03625492939829538, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 9.275972843170166e-07, "grad_norm": 7.2718987464904785, "learning_rate": 1.2038999576091563e-07, "loss": 0.6284, "mean_token_accuracy": 0.8102662563323975, "num_tokens": 10898800.0, "step": 285 }, { "epoch": 0.036382139676885895, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 2.115964889526367e-06, "ewc_loss_parallel": 9.462237358093262e-07, "grad_norm": 5.945026874542236, "learning_rate": 1.208139041966935e-07, "loss": 0.5935, "mean_token_accuracy": 0.8206892609596252, "num_tokens": 10937462.0, "step": 286 }, { "epoch": 0.0365093499554764, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 2.0712614059448242e-06, "ewc_loss_parallel": 8.977949619293213e-07, "grad_norm": 5.0751423835754395, "learning_rate": 1.2123781263247137e-07, "loss": 0.6546, "mean_token_accuracy": 0.798488199710846, "num_tokens": 10970739.0, "step": 287 }, { "epoch": 0.03663656023406691, "ewc_loss": 0.00106048583984375, "ewc_loss_diag": 2.0712614059448242e-06, "ewc_loss_parallel": 8.493661880493164e-07, "grad_norm": 5.691258907318115, "learning_rate": 1.2166172106824924e-07, "loss": 0.7069, "mean_token_accuracy": 0.7930759787559509, "num_tokens": 11003000.0, "step": 288 }, { "epoch": 0.036763770512657425, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 2.086162567138672e-06, "ewc_loss_parallel": 8.642673492431641e-07, "grad_norm": 7.1439642906188965, "learning_rate": 1.2208562950402712e-07, "loss": 0.6121, "mean_token_accuracy": 0.8105375170707703, "num_tokens": 11039665.0, "step": 289 }, { "epoch": 0.03689098079124793, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 2.1010637283325195e-06, "ewc_loss_parallel": 9.164214134216309e-07, "grad_norm": 6.7233147621154785, "learning_rate": 1.22509537939805e-07, "loss": 0.6402, "mean_token_accuracy": 0.8008694648742676, "num_tokens": 11078368.0, "step": 290 }, { "epoch": 0.03701819106983844, "ewc_loss": 0.0011444091796875, "ewc_loss_diag": 2.115964889526367e-06, "ewc_loss_parallel": 9.350478649139404e-07, "grad_norm": 4.751568794250488, "learning_rate": 1.2293344637558286e-07, "loss": 0.6651, "mean_token_accuracy": 0.7994508743286133, "num_tokens": 11122654.0, "step": 291 }, { "epoch": 0.037145401348428954, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 8.754432201385498e-07, "grad_norm": 5.508669853210449, "learning_rate": 1.2335735481136073e-07, "loss": 0.6197, "mean_token_accuracy": 0.8059749603271484, "num_tokens": 11162622.0, "step": 292 }, { "epoch": 0.03727261162701946, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 8.828938007354736e-07, "grad_norm": 5.822005271911621, "learning_rate": 1.237812632471386e-07, "loss": 0.6315, "mean_token_accuracy": 0.8086093664169312, "num_tokens": 11202564.0, "step": 293 }, { "epoch": 0.037399821905609976, "ewc_loss": 0.00112152099609375, "ewc_loss_diag": 2.16066837310791e-06, "ewc_loss_parallel": 9.08970832824707e-07, "grad_norm": 6.33915901184082, "learning_rate": 1.2420517168291648e-07, "loss": 0.6497, "mean_token_accuracy": 0.7996283173561096, "num_tokens": 11243422.0, "step": 294 }, { "epoch": 0.03752703218420048, "ewc_loss": 0.0011444091796875, "ewc_loss_diag": 2.175569534301758e-06, "ewc_loss_parallel": 9.313225746154785e-07, "grad_norm": 6.154113292694092, "learning_rate": 1.2462908011869435e-07, "loss": 0.6315, "mean_token_accuracy": 0.805094838142395, "num_tokens": 11280867.0, "step": 295 }, { "epoch": 0.03765424246279099, "ewc_loss": 0.00115203857421875, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.313225746154785e-07, "grad_norm": 14.11233901977539, "learning_rate": 1.2505298855447223e-07, "loss": 0.5807, "mean_token_accuracy": 0.8214888572692871, "num_tokens": 11318454.0, "step": 296 }, { "epoch": 0.037781452741381505, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 5.598068714141846, "learning_rate": 1.254768969902501e-07, "loss": 0.7209, "mean_token_accuracy": 0.783334493637085, "num_tokens": 11354721.0, "step": 297 }, { "epoch": 0.03790866301997201, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.142246246337891, "learning_rate": 1.2590080542602797e-07, "loss": 0.6786, "mean_token_accuracy": 0.7952332496643066, "num_tokens": 11385938.0, "step": 298 }, { "epoch": 0.03803587329856253, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 8.977949619293213e-07, "grad_norm": 5.481168270111084, "learning_rate": 1.2632471386180584e-07, "loss": 0.6584, "mean_token_accuracy": 0.8000199794769287, "num_tokens": 11424911.0, "step": 299 }, { "epoch": 0.038163083577153034, "ewc_loss": 0.00112152099609375, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.015202522277832e-07, "grad_norm": 5.411185264587402, "learning_rate": 1.2674862229758372e-07, "loss": 0.6262, "mean_token_accuracy": 0.8103656768798828, "num_tokens": 11460580.0, "step": 300 }, { "epoch": 0.03829029385574354, "ewc_loss": 0.00112152099609375, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.052455425262451e-07, "grad_norm": 5.49587345123291, "learning_rate": 1.271725307333616e-07, "loss": 0.6168, "mean_token_accuracy": 0.813427746295929, "num_tokens": 11501031.0, "step": 301 }, { "epoch": 0.038417504134334056, "ewc_loss": 0.00115203857421875, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.350478649139404e-07, "grad_norm": 5.633916854858398, "learning_rate": 1.2759643916913946e-07, "loss": 0.6264, "mean_token_accuracy": 0.8095259666442871, "num_tokens": 11544053.0, "step": 302 }, { "epoch": 0.03854471441292456, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 9.462237358093262e-07, "grad_norm": 5.91286563873291, "learning_rate": 1.2802034760491733e-07, "loss": 0.6422, "mean_token_accuracy": 0.8042821884155273, "num_tokens": 11576944.0, "step": 303 }, { "epoch": 0.03867192469151508, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 9.611248970031738e-07, "grad_norm": 4.755675315856934, "learning_rate": 1.284442560406952e-07, "loss": 0.6928, "mean_token_accuracy": 0.7894173860549927, "num_tokens": 11619735.0, "step": 304 }, { "epoch": 0.038799134970105585, "ewc_loss": 0.0011444091796875, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 9.238719940185547e-07, "grad_norm": 7.4217400550842285, "learning_rate": 1.2886816447647308e-07, "loss": 0.7297, "mean_token_accuracy": 0.7771246433258057, "num_tokens": 11655025.0, "step": 305 }, { "epoch": 0.03892634524869609, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 6.446855545043945, "learning_rate": 1.2929207291225095e-07, "loss": 0.6274, "mean_token_accuracy": 0.8090263605117798, "num_tokens": 11691970.0, "step": 306 }, { "epoch": 0.03905355552728661, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.294778823852539e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 5.85332727432251, "learning_rate": 1.2971598134802882e-07, "loss": 0.6429, "mean_token_accuracy": 0.8034597039222717, "num_tokens": 11729561.0, "step": 307 }, { "epoch": 0.039180765805877114, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.419530391693115, "learning_rate": 1.301398897838067e-07, "loss": 0.6618, "mean_token_accuracy": 0.797075629234314, "num_tokens": 11765707.0, "step": 308 }, { "epoch": 0.03930797608446762, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 2.250075340270996e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.515475749969482, "learning_rate": 1.3056379821958457e-07, "loss": 0.671, "mean_token_accuracy": 0.8016029596328735, "num_tokens": 11806223.0, "step": 309 }, { "epoch": 0.039435186363058136, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 2.2798776626586914e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 6.170906066894531, "learning_rate": 1.3098770665536244e-07, "loss": 0.6263, "mean_token_accuracy": 0.8108688592910767, "num_tokens": 11845477.0, "step": 310 }, { "epoch": 0.039562396641648644, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 2.2798776626586914e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 6.63601016998291, "learning_rate": 1.3141161509114031e-07, "loss": 0.6104, "mean_token_accuracy": 0.8093119263648987, "num_tokens": 11877973.0, "step": 311 }, { "epoch": 0.03968960692023916, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 2.2798776626586914e-06, "ewc_loss_parallel": 9.98377799987793e-07, "grad_norm": 5.7169270515441895, "learning_rate": 1.3183552352691819e-07, "loss": 0.6173, "mean_token_accuracy": 0.8120486736297607, "num_tokens": 11912093.0, "step": 312 }, { "epoch": 0.039816817198829665, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.294778823852539e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 5.745302200317383, "learning_rate": 1.3225943196269603e-07, "loss": 0.6426, "mean_token_accuracy": 0.8076736927032471, "num_tokens": 11952541.0, "step": 313 }, { "epoch": 0.03994402747742017, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.3096799850463867e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 5.881008148193359, "learning_rate": 1.3268334039847393e-07, "loss": 0.6201, "mean_token_accuracy": 0.8086205124855042, "num_tokens": 11990414.0, "step": 314 }, { "epoch": 0.04007123775601069, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.3096799850463867e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 5.244340419769287, "learning_rate": 1.3310724883425178e-07, "loss": 0.6878, "mean_token_accuracy": 0.7887126207351685, "num_tokens": 12025925.0, "step": 315 }, { "epoch": 0.040198448034601195, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.3245811462402344e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.365887641906738, "learning_rate": 1.3353115727002968e-07, "loss": 0.6545, "mean_token_accuracy": 0.8002618551254272, "num_tokens": 12068879.0, "step": 316 }, { "epoch": 0.04032565831319171, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.3245811462402344e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 4.765003681182861, "learning_rate": 1.3395506570580752e-07, "loss": 0.6208, "mean_token_accuracy": 0.8095247745513916, "num_tokens": 12111931.0, "step": 317 }, { "epoch": 0.040452868591782216, "ewc_loss": 0.001220703125, "ewc_loss_diag": 2.3543834686279297e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 6.428534984588623, "learning_rate": 1.3437897414158542e-07, "loss": 0.6893, "mean_token_accuracy": 0.7879332900047302, "num_tokens": 12140431.0, "step": 318 }, { "epoch": 0.040580078870372724, "ewc_loss": 0.001251220703125, "ewc_loss_diag": 2.3543834686279297e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 5.937829494476318, "learning_rate": 1.3480288257736327e-07, "loss": 0.6183, "mean_token_accuracy": 0.8094395399093628, "num_tokens": 12173932.0, "step": 319 }, { "epoch": 0.04070728914896324, "ewc_loss": 0.0012664794921875, "ewc_loss_diag": 2.384185791015625e-06, "ewc_loss_parallel": 1.0281801223754883e-06, "grad_norm": 5.2870402336120605, "learning_rate": 1.3522679101314117e-07, "loss": 0.5965, "mean_token_accuracy": 0.8143730163574219, "num_tokens": 12210040.0, "step": 320 }, { "epoch": 0.040834499427553746, "ewc_loss": 0.0012664794921875, "ewc_loss_diag": 2.384185791015625e-06, "ewc_loss_parallel": 1.0281801223754883e-06, "grad_norm": 6.036568641662598, "learning_rate": 1.35650699448919e-07, "loss": 0.6294, "mean_token_accuracy": 0.8088676333427429, "num_tokens": 12248556.0, "step": 321 }, { "epoch": 0.04096170970614425, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.4139881134033203e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 5.649049758911133, "learning_rate": 1.360746078846969e-07, "loss": 0.6183, "mean_token_accuracy": 0.8100505471229553, "num_tokens": 12285664.0, "step": 322 }, { "epoch": 0.04108891998473477, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.4139881134033203e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 6.41366720199585, "learning_rate": 1.3649851632047476e-07, "loss": 0.6717, "mean_token_accuracy": 0.7950977683067322, "num_tokens": 12326964.0, "step": 323 }, { "epoch": 0.041216130263325275, "ewc_loss": 0.00131988525390625, "ewc_loss_diag": 2.4139881134033203e-06, "ewc_loss_parallel": 1.0728836059570312e-06, "grad_norm": 5.382884502410889, "learning_rate": 1.3692242475625266e-07, "loss": 0.5801, "mean_token_accuracy": 0.8242741823196411, "num_tokens": 12366541.0, "step": 324 }, { "epoch": 0.04134334054191579, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.4139881134033203e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 6.8196539878845215, "learning_rate": 1.373463331920305e-07, "loss": 0.642, "mean_token_accuracy": 0.8090898394584656, "num_tokens": 12405664.0, "step": 325 }, { "epoch": 0.0414705508205063, "ewc_loss": 0.00133514404296875, "ewc_loss_diag": 2.428889274597168e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 6.298036098480225, "learning_rate": 1.377702416278084e-07, "loss": 0.7064, "mean_token_accuracy": 0.7860493063926697, "num_tokens": 12445039.0, "step": 326 }, { "epoch": 0.041597761099096804, "ewc_loss": 0.00133514404296875, "ewc_loss_diag": 2.428889274597168e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 5.238296985626221, "learning_rate": 1.3819415006358625e-07, "loss": 0.5763, "mean_token_accuracy": 0.8229102492332458, "num_tokens": 12485481.0, "step": 327 }, { "epoch": 0.04172497137768732, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 1.043081283569336e-06, "grad_norm": 5.591993808746338, "learning_rate": 1.3861805849936415e-07, "loss": 0.6013, "mean_token_accuracy": 0.8183708190917969, "num_tokens": 12530272.0, "step": 328 }, { "epoch": 0.041852181656277826, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 6.873601913452148, "learning_rate": 1.39041966935142e-07, "loss": 0.6473, "mean_token_accuracy": 0.801405668258667, "num_tokens": 12565854.0, "step": 329 }, { "epoch": 0.04197939193486834, "ewc_loss": 0.0013427734375, "ewc_loss_diag": 2.473592758178711e-06, "ewc_loss_parallel": 1.0952353477478027e-06, "grad_norm": 5.268452167510986, "learning_rate": 1.394658753709199e-07, "loss": 0.633, "mean_token_accuracy": 0.8068411350250244, "num_tokens": 12607313.0, "step": 330 }, { "epoch": 0.04210660221345885, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 2.473592758178711e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 5.46217679977417, "learning_rate": 1.3988978380669774e-07, "loss": 0.6707, "mean_token_accuracy": 0.7959358096122742, "num_tokens": 12642057.0, "step": 331 }, { "epoch": 0.042233812492049355, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 2.4884939193725586e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 5.335847854614258, "learning_rate": 1.403136922424756e-07, "loss": 0.6619, "mean_token_accuracy": 0.7973973751068115, "num_tokens": 12684118.0, "step": 332 }, { "epoch": 0.04236102277063987, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.4884939193725586e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 5.532853126525879, "learning_rate": 1.4073760067825348e-07, "loss": 0.6265, "mean_token_accuracy": 0.8074829578399658, "num_tokens": 12721945.0, "step": 333 }, { "epoch": 0.04248823304923038, "ewc_loss": 0.00130462646484375, "ewc_loss_diag": 2.518296241760254e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 5.569463729858398, "learning_rate": 1.4116150911403136e-07, "loss": 0.6411, "mean_token_accuracy": 0.8035536408424377, "num_tokens": 12764175.0, "step": 334 }, { "epoch": 0.04261544332782089, "ewc_loss": 0.00133514404296875, "ewc_loss_diag": 2.5033950805664062e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 5.83506965637207, "learning_rate": 1.4158541754980923e-07, "loss": 0.5913, "mean_token_accuracy": 0.8203176856040955, "num_tokens": 12801600.0, "step": 335 }, { "epoch": 0.0427426536064114, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.5033950805664062e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 6.464240550994873, "learning_rate": 1.420093259855871e-07, "loss": 0.6231, "mean_token_accuracy": 0.8158385157585144, "num_tokens": 12843584.0, "step": 336 }, { "epoch": 0.042869863885001906, "ewc_loss": 0.00138092041015625, "ewc_loss_diag": 2.518296241760254e-06, "ewc_loss_parallel": 1.1324882507324219e-06, "grad_norm": 5.564693927764893, "learning_rate": 1.4243323442136497e-07, "loss": 0.6369, "mean_token_accuracy": 0.8046844005584717, "num_tokens": 12878105.0, "step": 337 }, { "epoch": 0.04299707416359242, "ewc_loss": 0.00135040283203125, "ewc_loss_diag": 2.5331974029541016e-06, "ewc_loss_parallel": 1.0952353477478027e-06, "grad_norm": 5.604008674621582, "learning_rate": 1.4285714285714285e-07, "loss": 0.6448, "mean_token_accuracy": 0.8038008213043213, "num_tokens": 12911712.0, "step": 338 }, { "epoch": 0.04312428444218293, "ewc_loss": 0.0013427734375, "ewc_loss_diag": 2.5480985641479492e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 5.788086414337158, "learning_rate": 1.4328105129292072e-07, "loss": 0.6257, "mean_token_accuracy": 0.8103423714637756, "num_tokens": 12951303.0, "step": 339 }, { "epoch": 0.043251494720773435, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.5779008865356445e-06, "ewc_loss_parallel": 1.0952353477478027e-06, "grad_norm": 5.779592990875244, "learning_rate": 1.437049597286986e-07, "loss": 0.6125, "mean_token_accuracy": 0.8122570514678955, "num_tokens": 12990544.0, "step": 340 }, { "epoch": 0.04337870499936395, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.562999725341797e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 6.566515922546387, "learning_rate": 1.4412886816447646e-07, "loss": 0.6714, "mean_token_accuracy": 0.7956002950668335, "num_tokens": 13023766.0, "step": 341 }, { "epoch": 0.04350591527795446, "ewc_loss": 0.00139617919921875, "ewc_loss_diag": 2.5779008865356445e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 5.8627610206604, "learning_rate": 1.4455277660025434e-07, "loss": 0.5852, "mean_token_accuracy": 0.820827066898346, "num_tokens": 13056337.0, "step": 342 }, { "epoch": 0.04363312555654497, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.592802047729492e-06, "ewc_loss_parallel": 1.1175870895385742e-06, "grad_norm": 5.810250759124756, "learning_rate": 1.449766850360322e-07, "loss": 0.6436, "mean_token_accuracy": 0.8045777678489685, "num_tokens": 13097345.0, "step": 343 }, { "epoch": 0.04376033583513548, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 7.395123481750488, "learning_rate": 1.4540059347181008e-07, "loss": 0.6292, "mean_token_accuracy": 0.8080604076385498, "num_tokens": 13127332.0, "step": 344 }, { "epoch": 0.043887546113725986, "ewc_loss": 0.0014190673828125, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 4.936971187591553, "learning_rate": 1.4582450190758795e-07, "loss": 0.6579, "mean_token_accuracy": 0.7978484034538269, "num_tokens": 13170760.0, "step": 345 }, { "epoch": 0.0440147563923165, "ewc_loss": 0.00136566162109375, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 7.027812957763672, "learning_rate": 1.4624841034336583e-07, "loss": 0.6264, "mean_token_accuracy": 0.8106507658958435, "num_tokens": 13207061.0, "step": 346 }, { "epoch": 0.04414196667090701, "ewc_loss": 0.00140380859375, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 5.814101696014404, "learning_rate": 1.466723187791437e-07, "loss": 0.5956, "mean_token_accuracy": 0.817988395690918, "num_tokens": 13247903.0, "step": 347 }, { "epoch": 0.04426917694949752, "ewc_loss": 0.00140380859375, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 5.352396011352539, "learning_rate": 1.4709622721492157e-07, "loss": 0.6059, "mean_token_accuracy": 0.811024010181427, "num_tokens": 13287193.0, "step": 348 }, { "epoch": 0.04439638722808803, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 6.009947299957275, "learning_rate": 1.4752013565069942e-07, "loss": 0.61, "mean_token_accuracy": 0.813698947429657, "num_tokens": 13323112.0, "step": 349 }, { "epoch": 0.04452359750667854, "ewc_loss": 0.0013885498046875, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1324882507324219e-06, "grad_norm": 5.4341816902160645, "learning_rate": 1.4794404408647732e-07, "loss": 0.6057, "mean_token_accuracy": 0.8151636123657227, "num_tokens": 13359653.0, "step": 350 }, { "epoch": 0.04465080778526905, "ewc_loss": 0.0013885498046875, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.125037670135498e-06, "grad_norm": 5.344046115875244, "learning_rate": 1.4836795252225516e-07, "loss": 0.5764, "mean_token_accuracy": 0.8211256265640259, "num_tokens": 13400571.0, "step": 351 }, { "epoch": 0.04477801806385956, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 5.797962188720703, "learning_rate": 1.4879186095803306e-07, "loss": 0.5546, "mean_token_accuracy": 0.8241362571716309, "num_tokens": 13437656.0, "step": 352 }, { "epoch": 0.04490522834245007, "ewc_loss": 0.00140380859375, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 5.380159378051758, "learning_rate": 1.492157693938109e-07, "loss": 0.6136, "mean_token_accuracy": 0.8118833899497986, "num_tokens": 13474688.0, "step": 353 }, { "epoch": 0.04503243862104058, "ewc_loss": 0.00140380859375, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 4.748111724853516, "learning_rate": 1.496396778295888e-07, "loss": 0.5511, "mean_token_accuracy": 0.8285671472549438, "num_tokens": 13518256.0, "step": 354 }, { "epoch": 0.04515964889963109, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 5.803289413452148, "learning_rate": 1.5006358626536665e-07, "loss": 0.6379, "mean_token_accuracy": 0.8067647814750671, "num_tokens": 13554315.0, "step": 355 }, { "epoch": 0.0452868591782216, "ewc_loss": 0.00141143798828125, "ewc_loss_diag": 2.652406692504883e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 5.4617719650268555, "learning_rate": 1.5048749470114455e-07, "loss": 0.6594, "mean_token_accuracy": 0.7998653054237366, "num_tokens": 13594388.0, "step": 356 }, { "epoch": 0.04541406945681211, "ewc_loss": 0.001434326171875, "ewc_loss_diag": 2.652406692504883e-06, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 5.295895576477051, "learning_rate": 1.509114031369224e-07, "loss": 0.5847, "mean_token_accuracy": 0.817017674446106, "num_tokens": 13633704.0, "step": 357 }, { "epoch": 0.04554127973540262, "ewc_loss": 0.00142669677734375, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.46293306350708, "learning_rate": 1.513353115727003e-07, "loss": 0.6351, "mean_token_accuracy": 0.8018635511398315, "num_tokens": 13671183.0, "step": 358 }, { "epoch": 0.04566849001399313, "ewc_loss": 0.00142669677734375, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.230466842651367, "learning_rate": 1.5175922000847814e-07, "loss": 0.5896, "mean_token_accuracy": 0.8194122314453125, "num_tokens": 13709640.0, "step": 359 }, { "epoch": 0.04579570029258364, "ewc_loss": 0.00142669677734375, "ewc_loss_diag": 2.682209014892578e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.2420196533203125, "learning_rate": 1.5218312844425604e-07, "loss": 0.6125, "mean_token_accuracy": 0.8048927783966064, "num_tokens": 13751114.0, "step": 360 }, { "epoch": 0.045922910571174154, "ewc_loss": 0.001434326171875, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 5.443918228149414, "learning_rate": 1.526070368800339e-07, "loss": 0.5429, "mean_token_accuracy": 0.8267419338226318, "num_tokens": 13786448.0, "step": 361 }, { "epoch": 0.04605012084976466, "ewc_loss": 0.0014495849609375, "ewc_loss_diag": 2.6971101760864258e-06, "ewc_loss_parallel": 1.1846423149108887e-06, "grad_norm": 4.884997844696045, "learning_rate": 1.530309453158118e-07, "loss": 0.566, "mean_token_accuracy": 0.8289527297019958, "num_tokens": 13827663.0, "step": 362 }, { "epoch": 0.04617733112835517, "ewc_loss": 0.001434326171875, "ewc_loss_diag": 2.7120113372802734e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.079734802246094, "learning_rate": 1.5345485375158963e-07, "loss": 0.6447, "mean_token_accuracy": 0.8038116097450256, "num_tokens": 13866078.0, "step": 363 }, { "epoch": 0.04630454140694568, "ewc_loss": 0.0014495849609375, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 1.1771917343139648e-06, "grad_norm": 4.921173572540283, "learning_rate": 1.5387876218736753e-07, "loss": 0.5221, "mean_token_accuracy": 0.8358685970306396, "num_tokens": 13909767.0, "step": 364 }, { "epoch": 0.04643175168553619, "ewc_loss": 0.00144195556640625, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 5.36339807510376, "learning_rate": 1.5430267062314538e-07, "loss": 0.6306, "mean_token_accuracy": 0.8084250688552856, "num_tokens": 13948937.0, "step": 365 }, { "epoch": 0.0465589619641267, "ewc_loss": 0.00148773193359375, "ewc_loss_diag": 2.7567148208618164e-06, "ewc_loss_parallel": 1.214444637298584e-06, "grad_norm": 6.920723915100098, "learning_rate": 1.5472657905892328e-07, "loss": 0.6414, "mean_token_accuracy": 0.802915096282959, "num_tokens": 13984039.0, "step": 366 }, { "epoch": 0.04668617224271721, "ewc_loss": 0.00156402587890625, "ewc_loss_diag": 2.7567148208618164e-06, "ewc_loss_parallel": 1.2889504432678223e-06, "grad_norm": 5.859159469604492, "learning_rate": 1.5515048749470113e-07, "loss": 0.6239, "mean_token_accuracy": 0.807212769985199, "num_tokens": 14018162.0, "step": 367 }, { "epoch": 0.04681338252130772, "ewc_loss": 0.00153350830078125, "ewc_loss_diag": 2.7567148208618164e-06, "ewc_loss_parallel": 1.259148120880127e-06, "grad_norm": 4.732341766357422, "learning_rate": 1.55574395930479e-07, "loss": 0.6234, "mean_token_accuracy": 0.8075419664382935, "num_tokens": 14056493.0, "step": 368 }, { "epoch": 0.046940592799898234, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 1.1920928955078125e-06, "grad_norm": 7.075473785400391, "learning_rate": 1.5599830436625687e-07, "loss": 0.6171, "mean_token_accuracy": 0.8108574151992798, "num_tokens": 14097530.0, "step": 369 }, { "epoch": 0.04706780307848874, "ewc_loss": 0.00154876708984375, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 1.2740492820739746e-06, "grad_norm": 5.248097896575928, "learning_rate": 1.5642221280203474e-07, "loss": 0.6354, "mean_token_accuracy": 0.8092803955078125, "num_tokens": 14136240.0, "step": 370 }, { "epoch": 0.04719501335707925, "ewc_loss": 0.0015106201171875, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 1.2367963790893555e-06, "grad_norm": 5.143188953399658, "learning_rate": 1.5684612123781262e-07, "loss": 0.5464, "mean_token_accuracy": 0.8301560282707214, "num_tokens": 14171010.0, "step": 371 }, { "epoch": 0.04732222363566976, "ewc_loss": 0.0014801025390625, "ewc_loss_diag": 2.7865171432495117e-06, "ewc_loss_parallel": 1.1995434761047363e-06, "grad_norm": 5.05763578414917, "learning_rate": 1.572700296735905e-07, "loss": 0.6173, "mean_token_accuracy": 0.8164548277854919, "num_tokens": 14206985.0, "step": 372 }, { "epoch": 0.04744943391426027, "ewc_loss": 0.0014801025390625, "ewc_loss_diag": 2.7865171432495117e-06, "ewc_loss_parallel": 1.2069940567016602e-06, "grad_norm": 6.352434158325195, "learning_rate": 1.576939381093684e-07, "loss": 0.6159, "mean_token_accuracy": 0.8075531721115112, "num_tokens": 14236728.0, "step": 373 }, { "epoch": 0.047576644192850785, "ewc_loss": 0.001556396484375, "ewc_loss_diag": 2.8014183044433594e-06, "ewc_loss_parallel": 1.2740492820739746e-06, "grad_norm": 5.172815322875977, "learning_rate": 1.5811784654514623e-07, "loss": 0.5426, "mean_token_accuracy": 0.8331494331359863, "num_tokens": 14270758.0, "step": 374 }, { "epoch": 0.04770385447144129, "ewc_loss": 0.00151824951171875, "ewc_loss_diag": 2.8014183044433594e-06, "ewc_loss_parallel": 1.2367963790893555e-06, "grad_norm": 4.639526844024658, "learning_rate": 1.5854175498092413e-07, "loss": 0.6599, "mean_token_accuracy": 0.7953880429267883, "num_tokens": 14315002.0, "step": 375 }, { "epoch": 0.0478310647500318, "ewc_loss": 0.00147247314453125, "ewc_loss_diag": 2.8461217880249023e-06, "ewc_loss_parallel": 1.1920928955078125e-06, "grad_norm": 5.346570014953613, "learning_rate": 1.5896566341670198e-07, "loss": 0.6299, "mean_token_accuracy": 0.8027976751327515, "num_tokens": 14353385.0, "step": 376 }, { "epoch": 0.047958275028622314, "ewc_loss": 0.00152587890625, "ewc_loss_diag": 2.8461217880249023e-06, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 6.396955966949463, "learning_rate": 1.5938957185247988e-07, "loss": 0.6633, "mean_token_accuracy": 0.7930671572685242, "num_tokens": 14391702.0, "step": 377 }, { "epoch": 0.04808548530721282, "ewc_loss": 0.00159454345703125, "ewc_loss_diag": 2.8312206268310547e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 6.034430980682373, "learning_rate": 1.5981348028825772e-07, "loss": 0.6307, "mean_token_accuracy": 0.8058227300643921, "num_tokens": 14427876.0, "step": 378 }, { "epoch": 0.048212695585803336, "ewc_loss": 0.00159454345703125, "ewc_loss_diag": 2.8312206268310547e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 6.1093430519104, "learning_rate": 1.6023738872403562e-07, "loss": 0.7178, "mean_token_accuracy": 0.778471827507019, "num_tokens": 14460349.0, "step": 379 }, { "epoch": 0.048339905864393844, "ewc_loss": 0.0015716552734375, "ewc_loss_diag": 2.86102294921875e-06, "ewc_loss_parallel": 1.2889504432678223e-06, "grad_norm": 5.1515350341796875, "learning_rate": 1.6066129715981347e-07, "loss": 0.5675, "mean_token_accuracy": 0.8262311816215515, "num_tokens": 14497763.0, "step": 380 }, { "epoch": 0.04846711614298435, "ewc_loss": 0.00153350830078125, "ewc_loss_diag": 2.8759241104125977e-06, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 4.781297206878662, "learning_rate": 1.6108520559559137e-07, "loss": 0.5609, "mean_token_accuracy": 0.8240915536880493, "num_tokens": 14537172.0, "step": 381 }, { "epoch": 0.048594326421574865, "ewc_loss": 0.00151824951171875, "ewc_loss_diag": 2.8759241104125977e-06, "ewc_loss_parallel": 1.2293457984924316e-06, "grad_norm": 6.769291877746582, "learning_rate": 1.6150911403136921e-07, "loss": 0.6485, "mean_token_accuracy": 0.8064388036727905, "num_tokens": 14565515.0, "step": 382 }, { "epoch": 0.04872153670016537, "ewc_loss": 0.00160980224609375, "ewc_loss_diag": 2.905726432800293e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 4.971029281616211, "learning_rate": 1.619330224671471e-07, "loss": 0.639, "mean_token_accuracy": 0.8051952123641968, "num_tokens": 14608104.0, "step": 383 }, { "epoch": 0.04884874697875588, "ewc_loss": 0.0015716552734375, "ewc_loss_diag": 2.9355287551879883e-06, "ewc_loss_parallel": 1.2740492820739746e-06, "grad_norm": 6.346306324005127, "learning_rate": 1.6235693090292496e-07, "loss": 0.6471, "mean_token_accuracy": 0.8049756288528442, "num_tokens": 14645328.0, "step": 384 }, { "epoch": 0.048975957257346395, "ewc_loss": 0.0016021728515625, "ewc_loss_diag": 2.9355287551879883e-06, "ewc_loss_parallel": 1.30385160446167e-06, "grad_norm": 5.680726051330566, "learning_rate": 1.6278083933870286e-07, "loss": 0.6809, "mean_token_accuracy": 0.7886072397232056, "num_tokens": 14678792.0, "step": 385 }, { "epoch": 0.0491031675359369, "ewc_loss": 0.0015869140625, "ewc_loss_diag": 2.9355287551879883e-06, "ewc_loss_parallel": 1.2889504432678223e-06, "grad_norm": 5.25646448135376, "learning_rate": 1.632047477744807e-07, "loss": 0.6035, "mean_token_accuracy": 0.8155178427696228, "num_tokens": 14715095.0, "step": 386 }, { "epoch": 0.049230377814527417, "ewc_loss": 0.00157928466796875, "ewc_loss_diag": 2.950429916381836e-06, "ewc_loss_parallel": 1.2814998626708984e-06, "grad_norm": 5.521219730377197, "learning_rate": 1.6362865621025858e-07, "loss": 0.6232, "mean_token_accuracy": 0.8094003200531006, "num_tokens": 14753641.0, "step": 387 }, { "epoch": 0.049357588093117924, "ewc_loss": 0.0015869140625, "ewc_loss_diag": 2.9802322387695312e-06, "ewc_loss_parallel": 1.2889504432678223e-06, "grad_norm": 5.1882548332214355, "learning_rate": 1.6405256464603645e-07, "loss": 0.5943, "mean_token_accuracy": 0.8151465654373169, "num_tokens": 14788816.0, "step": 388 }, { "epoch": 0.04948479837170843, "ewc_loss": 0.00156402587890625, "ewc_loss_diag": 2.9802322387695312e-06, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 5.66002893447876, "learning_rate": 1.6447647308181432e-07, "loss": 0.6381, "mean_token_accuracy": 0.803217887878418, "num_tokens": 14821607.0, "step": 389 }, { "epoch": 0.049612008650298946, "ewc_loss": 0.00159454345703125, "ewc_loss_diag": 2.9802322387695312e-06, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 4.706424713134766, "learning_rate": 1.649003815175922e-07, "loss": 0.6557, "mean_token_accuracy": 0.8008813858032227, "num_tokens": 14861298.0, "step": 390 }, { "epoch": 0.04973921892888945, "ewc_loss": 0.001556396484375, "ewc_loss_diag": 2.9802322387695312e-06, "ewc_loss_parallel": 1.259148120880127e-06, "grad_norm": 4.717723369598389, "learning_rate": 1.6532428995337007e-07, "loss": 0.5727, "mean_token_accuracy": 0.8216297030448914, "num_tokens": 14899951.0, "step": 391 }, { "epoch": 0.04986642920747997, "ewc_loss": 0.001556396484375, "ewc_loss_diag": 2.9802322387695312e-06, "ewc_loss_parallel": 1.259148120880127e-06, "grad_norm": 4.725744724273682, "learning_rate": 1.6574819838914794e-07, "loss": 0.5528, "mean_token_accuracy": 0.8293534517288208, "num_tokens": 14937169.0, "step": 392 }, { "epoch": 0.049993639486070475, "ewc_loss": 0.0015716552734375, "ewc_loss_diag": 2.995133399963379e-06, "ewc_loss_parallel": 1.2740492820739746e-06, "grad_norm": 5.064987659454346, "learning_rate": 1.661721068249258e-07, "loss": 0.6055, "mean_token_accuracy": 0.8129748106002808, "num_tokens": 14975549.0, "step": 393 }, { "epoch": 0.05012084976466098, "ewc_loss": 0.0016021728515625, "ewc_loss_diag": 3.0249357223510742e-06, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 5.408459663391113, "learning_rate": 1.6659601526070368e-07, "loss": 0.6335, "mean_token_accuracy": 0.805917501449585, "num_tokens": 15016630.0, "step": 394 }, { "epoch": 0.0502480600432515, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 3.0249357223510742e-06, "ewc_loss_parallel": 1.3262033462524414e-06, "grad_norm": 5.067709922790527, "learning_rate": 1.6701992369648156e-07, "loss": 0.615, "mean_token_accuracy": 0.8095262050628662, "num_tokens": 15057822.0, "step": 395 }, { "epoch": 0.050375270321842004, "ewc_loss": 0.00162506103515625, "ewc_loss_diag": 3.0249357223510742e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 5.956920623779297, "learning_rate": 1.6744383213225943e-07, "loss": 0.5882, "mean_token_accuracy": 0.8134816884994507, "num_tokens": 15094385.0, "step": 396 }, { "epoch": 0.05050248060043251, "ewc_loss": 0.00167083740234375, "ewc_loss_diag": 3.0249357223510742e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 5.09173059463501, "learning_rate": 1.678677405680373e-07, "loss": 0.5477, "mean_token_accuracy": 0.8267655372619629, "num_tokens": 15130969.0, "step": 397 }, { "epoch": 0.050629690879023026, "ewc_loss": 0.00162506103515625, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 6.430166721343994, "learning_rate": 1.6829164900381518e-07, "loss": 0.5691, "mean_token_accuracy": 0.8186612129211426, "num_tokens": 15166586.0, "step": 398 }, { "epoch": 0.05075690115761353, "ewc_loss": 0.00170135498046875, "ewc_loss_diag": 3.0547380447387695e-06, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 5.08831787109375, "learning_rate": 1.6871555743959305e-07, "loss": 0.5592, "mean_token_accuracy": 0.8227459192276001, "num_tokens": 15209603.0, "step": 399 }, { "epoch": 0.05088411143620405, "ewc_loss": 0.001617431640625, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 4.881830215454102, "learning_rate": 1.6913946587537092e-07, "loss": 0.5759, "mean_token_accuracy": 0.8189985752105713, "num_tokens": 15251542.0, "step": 400 }, { "epoch": 0.051011321714794555, "ewc_loss": 0.0016021728515625, "ewc_loss_diag": 3.084540367126465e-06, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 6.775312900543213, "learning_rate": 1.695633743111488e-07, "loss": 0.5911, "mean_token_accuracy": 0.8180962204933167, "num_tokens": 15288940.0, "step": 401 }, { "epoch": 0.05113853199338506, "ewc_loss": 0.00171661376953125, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 1.4081597328186035e-06, "grad_norm": 5.687469482421875, "learning_rate": 1.6998728274692667e-07, "loss": 0.558, "mean_token_accuracy": 0.829688549041748, "num_tokens": 15321965.0, "step": 402 }, { "epoch": 0.05126574227197558, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 1.385807991027832e-06, "grad_norm": 5.154510974884033, "learning_rate": 1.7041119118270454e-07, "loss": 0.566, "mean_token_accuracy": 0.8252189755439758, "num_tokens": 15362895.0, "step": 403 }, { "epoch": 0.051392952550566084, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 6.491892337799072, "learning_rate": 1.7083509961848238e-07, "loss": 0.6247, "mean_token_accuracy": 0.8058133721351624, "num_tokens": 15396675.0, "step": 404 }, { "epoch": 0.0515201628291566, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 1.3709068298339844e-06, "grad_norm": 5.73973274230957, "learning_rate": 1.7125900805426028e-07, "loss": 0.6454, "mean_token_accuracy": 0.8021761178970337, "num_tokens": 15443070.0, "step": 405 }, { "epoch": 0.051647373107747106, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.084540367126465e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 5.271401405334473, "learning_rate": 1.7168291649003813e-07, "loss": 0.5729, "mean_token_accuracy": 0.8211647868156433, "num_tokens": 15476747.0, "step": 406 }, { "epoch": 0.051774583386337614, "ewc_loss": 0.00162506103515625, "ewc_loss_diag": 3.11434268951416e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 5.389003753662109, "learning_rate": 1.7210682492581603e-07, "loss": 0.5983, "mean_token_accuracy": 0.8158057332038879, "num_tokens": 15514866.0, "step": 407 }, { "epoch": 0.05190179366492813, "ewc_loss": 0.001617431640625, "ewc_loss_diag": 3.1441450119018555e-06, "ewc_loss_parallel": 1.30385160446167e-06, "grad_norm": 5.278901100158691, "learning_rate": 1.7253073336159387e-07, "loss": 0.5752, "mean_token_accuracy": 0.8190435171127319, "num_tokens": 15554245.0, "step": 408 }, { "epoch": 0.052029003943518635, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 5.381374835968018, "learning_rate": 1.7295464179737177e-07, "loss": 0.6044, "mean_token_accuracy": 0.8139225244522095, "num_tokens": 15593958.0, "step": 409 }, { "epoch": 0.05215621422210915, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.3262033462524414e-06, "grad_norm": 5.259411811828613, "learning_rate": 1.7337855023314962e-07, "loss": 0.5489, "mean_token_accuracy": 0.8278770446777344, "num_tokens": 15632567.0, "step": 410 }, { "epoch": 0.05228342450069966, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.1739473342895508e-06, "ewc_loss_parallel": 1.3262033462524414e-06, "grad_norm": 5.686230182647705, "learning_rate": 1.7380245866892752e-07, "loss": 0.5584, "mean_token_accuracy": 0.8242603540420532, "num_tokens": 15668255.0, "step": 411 }, { "epoch": 0.052410634779290165, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.1739473342895508e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 4.921359539031982, "learning_rate": 1.7422636710470536e-07, "loss": 0.6113, "mean_token_accuracy": 0.8092986345291138, "num_tokens": 15705435.0, "step": 412 }, { "epoch": 0.05253784505788068, "ewc_loss": 0.001617431640625, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 4.967209815979004, "learning_rate": 1.7465027554048326e-07, "loss": 0.6314, "mean_token_accuracy": 0.8044681549072266, "num_tokens": 15744657.0, "step": 413 }, { "epoch": 0.05266505533647119, "ewc_loss": 0.001617431640625, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.30385160446167e-06, "grad_norm": 5.337074279785156, "learning_rate": 1.750741839762611e-07, "loss": 0.5748, "mean_token_accuracy": 0.822704553604126, "num_tokens": 15777374.0, "step": 414 }, { "epoch": 0.052792265615061694, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 4.983302116394043, "learning_rate": 1.75498092412039e-07, "loss": 0.5468, "mean_token_accuracy": 0.8297377228736877, "num_tokens": 15814047.0, "step": 415 }, { "epoch": 0.05291947589365221, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.1888484954833984e-06, "ewc_loss_parallel": 1.3262033462524414e-06, "grad_norm": 4.863253116607666, "learning_rate": 1.7592200084781686e-07, "loss": 0.6286, "mean_token_accuracy": 0.8053648471832275, "num_tokens": 15849110.0, "step": 416 }, { "epoch": 0.053046686172242716, "ewc_loss": 0.0016632080078125, "ewc_loss_diag": 3.2186508178710938e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 5.2371954917907715, "learning_rate": 1.7634590928359475e-07, "loss": 0.5988, "mean_token_accuracy": 0.8167574405670166, "num_tokens": 15888910.0, "step": 417 }, { "epoch": 0.05317389645083323, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.248453140258789e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 5.006865978240967, "learning_rate": 1.767698177193726e-07, "loss": 0.5851, "mean_token_accuracy": 0.8179670572280884, "num_tokens": 15925604.0, "step": 418 }, { "epoch": 0.05330110672942374, "ewc_loss": 0.00168609619140625, "ewc_loss_diag": 3.2633543014526367e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 4.727658271789551, "learning_rate": 1.771937261551505e-07, "loss": 0.5587, "mean_token_accuracy": 0.8323838710784912, "num_tokens": 15961401.0, "step": 419 }, { "epoch": 0.053428317008014245, "ewc_loss": 0.00167083740234375, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 5.5727925300598145, "learning_rate": 1.7761763459092835e-07, "loss": 0.59, "mean_token_accuracy": 0.8172739744186401, "num_tokens": 16001029.0, "step": 420 }, { "epoch": 0.05355552728660476, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.4081597328186035e-06, "grad_norm": 5.34824800491333, "learning_rate": 1.7804154302670624e-07, "loss": 0.5711, "mean_token_accuracy": 0.8227994441986084, "num_tokens": 16036014.0, "step": 421 }, { "epoch": 0.05368273756519527, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.4081597328186035e-06, "grad_norm": 4.968070983886719, "learning_rate": 1.784654514624841e-07, "loss": 0.5526, "mean_token_accuracy": 0.82462477684021, "num_tokens": 16075373.0, "step": 422 }, { "epoch": 0.05380994784378578, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 4.923968315124512, "learning_rate": 1.7888935989826196e-07, "loss": 0.5948, "mean_token_accuracy": 0.8114888072013855, "num_tokens": 16111352.0, "step": 423 }, { "epoch": 0.05393715812237629, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 5.341318607330322, "learning_rate": 1.7931326833403984e-07, "loss": 0.6114, "mean_token_accuracy": 0.8091445565223694, "num_tokens": 16148826.0, "step": 424 }, { "epoch": 0.054064368400966796, "ewc_loss": 0.00174713134765625, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.4156103134155273e-06, "grad_norm": 4.885394096374512, "learning_rate": 1.797371767698177e-07, "loss": 0.5186, "mean_token_accuracy": 0.8388264179229736, "num_tokens": 16193491.0, "step": 425 }, { "epoch": 0.05419157867955731, "ewc_loss": 0.001708984375, "ewc_loss_diag": 3.3229589462280273e-06, "ewc_loss_parallel": 1.3709068298339844e-06, "grad_norm": 5.139915943145752, "learning_rate": 1.8016108520559558e-07, "loss": 0.5681, "mean_token_accuracy": 0.8240527510643005, "num_tokens": 16235595.0, "step": 426 }, { "epoch": 0.05431878895814782, "ewc_loss": 0.001708984375, "ewc_loss_diag": 3.3229589462280273e-06, "ewc_loss_parallel": 1.3783574104309082e-06, "grad_norm": 5.077657222747803, "learning_rate": 1.8058499364137345e-07, "loss": 0.622, "mean_token_accuracy": 0.8118182420730591, "num_tokens": 16274105.0, "step": 427 }, { "epoch": 0.054445999236738325, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.337860107421875e-06, "ewc_loss_parallel": 1.4007091522216797e-06, "grad_norm": 4.89215612411499, "learning_rate": 1.8100890207715133e-07, "loss": 0.6654, "mean_token_accuracy": 0.7910330295562744, "num_tokens": 16315886.0, "step": 428 }, { "epoch": 0.05457320951532884, "ewc_loss": 0.00173187255859375, "ewc_loss_diag": 3.337860107421875e-06, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 5.39816951751709, "learning_rate": 1.814328105129292e-07, "loss": 0.6069, "mean_token_accuracy": 0.8106334805488586, "num_tokens": 16353021.0, "step": 429 }, { "epoch": 0.05470041979391935, "ewc_loss": 0.0017852783203125, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 1.4528632164001465e-06, "grad_norm": 6.146572113037109, "learning_rate": 1.8185671894870707e-07, "loss": 0.5821, "mean_token_accuracy": 0.8158260583877563, "num_tokens": 16380887.0, "step": 430 }, { "epoch": 0.05482763007250986, "ewc_loss": 0.00189208984375, "ewc_loss_diag": 3.3676624298095703e-06, "ewc_loss_parallel": 1.55717134475708e-06, "grad_norm": 4.922359466552734, "learning_rate": 1.8228062738448494e-07, "loss": 0.6377, "mean_token_accuracy": 0.8042577505111694, "num_tokens": 16421066.0, "step": 431 }, { "epoch": 0.05495484035110037, "ewc_loss": 0.00177001953125, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.4230608940124512e-06, "grad_norm": 5.093000888824463, "learning_rate": 1.8270453582026282e-07, "loss": 0.606, "mean_token_accuracy": 0.8111566305160522, "num_tokens": 16457197.0, "step": 432 }, { "epoch": 0.055082050629690876, "ewc_loss": 0.00174713134765625, "ewc_loss_diag": 3.3676624298095703e-06, "ewc_loss_parallel": 1.4081597328186035e-06, "grad_norm": 4.669362545013428, "learning_rate": 1.831284442560407e-07, "loss": 0.5609, "mean_token_accuracy": 0.8243941068649292, "num_tokens": 16497385.0, "step": 433 }, { "epoch": 0.05520926090828139, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.382563591003418e-06, "ewc_loss_parallel": 1.4007091522216797e-06, "grad_norm": 5.110506534576416, "learning_rate": 1.8355235269181856e-07, "loss": 0.5762, "mean_token_accuracy": 0.8211747407913208, "num_tokens": 16540099.0, "step": 434 }, { "epoch": 0.0553364711868719, "ewc_loss": 0.00179290771484375, "ewc_loss_diag": 3.382563591003418e-06, "ewc_loss_parallel": 1.4603137969970703e-06, "grad_norm": 5.4074788093566895, "learning_rate": 1.8397626112759643e-07, "loss": 0.5734, "mean_token_accuracy": 0.8207811713218689, "num_tokens": 16572760.0, "step": 435 }, { "epoch": 0.05546368146546241, "ewc_loss": 0.0018310546875, "ewc_loss_diag": 3.3974647521972656e-06, "ewc_loss_parallel": 1.4975666999816895e-06, "grad_norm": 5.007583141326904, "learning_rate": 1.844001695633743e-07, "loss": 0.5325, "mean_token_accuracy": 0.8320367336273193, "num_tokens": 16608549.0, "step": 436 }, { "epoch": 0.05559089174405292, "ewc_loss": 0.00179290771484375, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.4528632164001465e-06, "grad_norm": 5.388584136962891, "learning_rate": 1.8482407799915218e-07, "loss": 0.5703, "mean_token_accuracy": 0.823280930519104, "num_tokens": 16643800.0, "step": 437 }, { "epoch": 0.05571810202264343, "ewc_loss": 0.0018310546875, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.4901161193847656e-06, "grad_norm": 5.845646381378174, "learning_rate": 1.8524798643493005e-07, "loss": 0.6435, "mean_token_accuracy": 0.7983892560005188, "num_tokens": 16677948.0, "step": 438 }, { "epoch": 0.05584531230123394, "ewc_loss": 0.00189208984375, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.5497207641601562e-06, "grad_norm": 4.932487487792969, "learning_rate": 1.8567189487070792e-07, "loss": 0.616, "mean_token_accuracy": 0.810122013092041, "num_tokens": 16716905.0, "step": 439 }, { "epoch": 0.05597252257982445, "ewc_loss": 0.00179290771484375, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.4528632164001465e-06, "grad_norm": 4.812097549438477, "learning_rate": 1.8609580330648577e-07, "loss": 0.5273, "mean_token_accuracy": 0.8334239721298218, "num_tokens": 16756158.0, "step": 440 }, { "epoch": 0.05609973285841496, "ewc_loss": 0.00177001953125, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.4230608940124512e-06, "grad_norm": 5.437073230743408, "learning_rate": 1.8651971174226367e-07, "loss": 0.619, "mean_token_accuracy": 0.8115503191947937, "num_tokens": 16794261.0, "step": 441 }, { "epoch": 0.05622694313700547, "ewc_loss": 0.0018463134765625, "ewc_loss_diag": 3.427267074584961e-06, "ewc_loss_parallel": 1.5050172805786133e-06, "grad_norm": 4.817844390869141, "learning_rate": 1.8694362017804152e-07, "loss": 0.5756, "mean_token_accuracy": 0.8191814422607422, "num_tokens": 16834953.0, "step": 442 }, { "epoch": 0.05635415341559598, "ewc_loss": 0.00182342529296875, "ewc_loss_diag": 3.427267074584961e-06, "ewc_loss_parallel": 1.4826655387878418e-06, "grad_norm": 4.753342151641846, "learning_rate": 1.8736752861381941e-07, "loss": 0.5344, "mean_token_accuracy": 0.8343391418457031, "num_tokens": 16873954.0, "step": 443 }, { "epoch": 0.05648136369418649, "ewc_loss": 0.00177764892578125, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 1.430511474609375e-06, "grad_norm": 5.216651916503906, "learning_rate": 1.8779143704959726e-07, "loss": 0.5574, "mean_token_accuracy": 0.8228045105934143, "num_tokens": 16906018.0, "step": 444 }, { "epoch": 0.056608573972777, "ewc_loss": 0.0018463134765625, "ewc_loss_diag": 3.427267074584961e-06, "ewc_loss_parallel": 1.5050172805786133e-06, "grad_norm": 5.237789154052734, "learning_rate": 1.8821534548537516e-07, "loss": 0.6106, "mean_token_accuracy": 0.8118156790733337, "num_tokens": 16938945.0, "step": 445 }, { "epoch": 0.05673578425136751, "ewc_loss": 0.0018768310546875, "ewc_loss_diag": 3.471970558166504e-06, "ewc_loss_parallel": 1.5348196029663086e-06, "grad_norm": 4.7687153816223145, "learning_rate": 1.88639253921153e-07, "loss": 0.5688, "mean_token_accuracy": 0.8246365785598755, "num_tokens": 16982214.0, "step": 446 }, { "epoch": 0.05686299452995802, "ewc_loss": 0.00179290771484375, "ewc_loss_diag": 3.5017728805541992e-06, "ewc_loss_parallel": 1.4454126358032227e-06, "grad_norm": 4.873961925506592, "learning_rate": 1.890631623569309e-07, "loss": 0.5768, "mean_token_accuracy": 0.8203224539756775, "num_tokens": 17024367.0, "step": 447 }, { "epoch": 0.05699020480854853, "ewc_loss": 0.0018310546875, "ewc_loss_diag": 3.516674041748047e-06, "ewc_loss_parallel": 1.4826655387878418e-06, "grad_norm": 5.312644004821777, "learning_rate": 1.8948707079270875e-07, "loss": 0.5995, "mean_token_accuracy": 0.8092917203903198, "num_tokens": 17057451.0, "step": 448 }, { "epoch": 0.057117415087139044, "ewc_loss": 0.0019073486328125, "ewc_loss_diag": 3.516674041748047e-06, "ewc_loss_parallel": 1.55717134475708e-06, "grad_norm": 6.108102321624756, "learning_rate": 1.8991097922848665e-07, "loss": 0.5513, "mean_token_accuracy": 0.8262833952903748, "num_tokens": 17094633.0, "step": 449 }, { "epoch": 0.05724462536572955, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.56137752532959e-06, "ewc_loss_parallel": 1.6167759895324707e-06, "grad_norm": 5.6215691566467285, "learning_rate": 1.903348876642645e-07, "loss": 0.6057, "mean_token_accuracy": 0.8093413710594177, "num_tokens": 17135892.0, "step": 450 }, { "epoch": 0.05737183564432006, "ewc_loss": 0.001922607421875, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 6.049203395843506, "learning_rate": 1.907587961000424e-07, "loss": 0.5485, "mean_token_accuracy": 0.8288309574127197, "num_tokens": 17173881.0, "step": 451 }, { "epoch": 0.05749904592291057, "ewc_loss": 0.00194549560546875, "ewc_loss_diag": 3.56137752532959e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 5.265893936157227, "learning_rate": 1.9118270453582024e-07, "loss": 0.5975, "mean_token_accuracy": 0.8153139352798462, "num_tokens": 17215668.0, "step": 452 }, { "epoch": 0.05762625620150108, "ewc_loss": 0.001861572265625, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.5124678611755371e-06, "grad_norm": 5.5880889892578125, "learning_rate": 1.9160661297159814e-07, "loss": 0.6214, "mean_token_accuracy": 0.8096848130226135, "num_tokens": 17254339.0, "step": 453 }, { "epoch": 0.057753466480091595, "ewc_loss": 0.00189208984375, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.5348196029663086e-06, "grad_norm": 5.340368270874023, "learning_rate": 1.9203052140737599e-07, "loss": 0.606, "mean_token_accuracy": 0.8092540502548218, "num_tokens": 17288858.0, "step": 454 }, { "epoch": 0.0578806767586821, "ewc_loss": 0.0018768310546875, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.5273690223693848e-06, "grad_norm": 4.840202331542969, "learning_rate": 1.9245442984315389e-07, "loss": 0.5823, "mean_token_accuracy": 0.8206375241279602, "num_tokens": 17330601.0, "step": 455 }, { "epoch": 0.05800788703727261, "ewc_loss": 0.0018463134765625, "ewc_loss_diag": 3.5762786865234375e-06, "ewc_loss_parallel": 1.4901161193847656e-06, "grad_norm": 5.130904197692871, "learning_rate": 1.9287833827893173e-07, "loss": 0.6016, "mean_token_accuracy": 0.8115954399108887, "num_tokens": 17367797.0, "step": 456 }, { "epoch": 0.058135097315863124, "ewc_loss": 0.00186920166015625, "ewc_loss_diag": 3.591179847717285e-06, "ewc_loss_parallel": 1.5124678611755371e-06, "grad_norm": 5.199310779571533, "learning_rate": 1.9330224671470963e-07, "loss": 0.5879, "mean_token_accuracy": 0.8168240785598755, "num_tokens": 17403206.0, "step": 457 }, { "epoch": 0.05826230759445363, "ewc_loss": 0.00188446044921875, "ewc_loss_diag": 3.591179847717285e-06, "ewc_loss_parallel": 1.5273690223693848e-06, "grad_norm": 5.134757041931152, "learning_rate": 1.9372615515048748e-07, "loss": 0.6335, "mean_token_accuracy": 0.8070715069770813, "num_tokens": 17437207.0, "step": 458 }, { "epoch": 0.05838951787304414, "ewc_loss": 0.00188446044921875, "ewc_loss_diag": 3.591179847717285e-06, "ewc_loss_parallel": 1.5273690223693848e-06, "grad_norm": 4.977890491485596, "learning_rate": 1.9415006358626535e-07, "loss": 0.5697, "mean_token_accuracy": 0.8247290849685669, "num_tokens": 17473335.0, "step": 459 }, { "epoch": 0.05851672815163465, "ewc_loss": 0.00189208984375, "ewc_loss_diag": 3.6209821701049805e-06, "ewc_loss_parallel": 1.5273690223693848e-06, "grad_norm": 5.415433406829834, "learning_rate": 1.9457397202204322e-07, "loss": 0.5598, "mean_token_accuracy": 0.8217307329177856, "num_tokens": 17505104.0, "step": 460 }, { "epoch": 0.05864393843022516, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.6209821701049805e-06, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 5.301265716552734, "learning_rate": 1.949978804578211e-07, "loss": 0.5439, "mean_token_accuracy": 0.8317750692367554, "num_tokens": 17539052.0, "step": 461 }, { "epoch": 0.058771148708815675, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.6209821701049805e-06, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 4.952563762664795, "learning_rate": 1.9542178889359897e-07, "loss": 0.6172, "mean_token_accuracy": 0.807295024394989, "num_tokens": 17578942.0, "step": 462 }, { "epoch": 0.05889835898740618, "ewc_loss": 0.001922607421875, "ewc_loss_diag": 3.6209821701049805e-06, "ewc_loss_parallel": 1.564621925354004e-06, "grad_norm": 5.2674031257629395, "learning_rate": 1.9584569732937684e-07, "loss": 0.5957, "mean_token_accuracy": 0.817896842956543, "num_tokens": 17611931.0, "step": 463 }, { "epoch": 0.05902556926599669, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.6209821701049805e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 5.248952865600586, "learning_rate": 1.962696057651547e-07, "loss": 0.5009, "mean_token_accuracy": 0.842167854309082, "num_tokens": 17648764.0, "step": 464 }, { "epoch": 0.059152779544587204, "ewc_loss": 0.00194549560546875, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 5.588404655456543, "learning_rate": 1.9669351420093258e-07, "loss": 0.6118, "mean_token_accuracy": 0.8094658851623535, "num_tokens": 17683099.0, "step": 465 }, { "epoch": 0.05927998982317771, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.6507844924926758e-06, "ewc_loss_parallel": 1.601874828338623e-06, "grad_norm": 5.0474162101745605, "learning_rate": 1.9711742263671046e-07, "loss": 0.5966, "mean_token_accuracy": 0.8185986876487732, "num_tokens": 17727120.0, "step": 466 }, { "epoch": 0.059407200101768226, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.680586814880371e-06, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 5.729621410369873, "learning_rate": 1.9754133107248833e-07, "loss": 0.5515, "mean_token_accuracy": 0.8216081261634827, "num_tokens": 17761676.0, "step": 467 }, { "epoch": 0.059534410380358734, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.680586814880371e-06, "ewc_loss_parallel": 1.6167759895324707e-06, "grad_norm": 5.326882839202881, "learning_rate": 1.979652395082662e-07, "loss": 0.6188, "mean_token_accuracy": 0.8093019723892212, "num_tokens": 17802311.0, "step": 468 }, { "epoch": 0.05966162065894924, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.6656856536865234e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 4.758208274841309, "learning_rate": 1.9838914794404408e-07, "loss": 0.5102, "mean_token_accuracy": 0.8364948034286499, "num_tokens": 17839048.0, "step": 469 }, { "epoch": 0.059788830937539755, "ewc_loss": 0.00189208984375, "ewc_loss_diag": 3.6656856536865234e-06, "ewc_loss_parallel": 1.5273690223693848e-06, "grad_norm": 4.656158924102783, "learning_rate": 1.9881305637982195e-07, "loss": 0.5816, "mean_token_accuracy": 0.8188114166259766, "num_tokens": 17881510.0, "step": 470 }, { "epoch": 0.05991604121613026, "ewc_loss": 0.0019073486328125, "ewc_loss_diag": 3.680586814880371e-06, "ewc_loss_parallel": 1.5422701835632324e-06, "grad_norm": 5.380643844604492, "learning_rate": 1.9923696481559982e-07, "loss": 0.6068, "mean_token_accuracy": 0.8137407302856445, "num_tokens": 17915080.0, "step": 471 }, { "epoch": 0.06004325149472077, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.6954879760742188e-06, "ewc_loss_parallel": 1.6167759895324707e-06, "grad_norm": 5.081576347351074, "learning_rate": 1.996608732513777e-07, "loss": 0.6395, "mean_token_accuracy": 0.8031994104385376, "num_tokens": 17957972.0, "step": 472 }, { "epoch": 0.060170461773311285, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.6954879760742188e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 4.753395080566406, "learning_rate": 2.0008478168715557e-07, "loss": 0.5105, "mean_token_accuracy": 0.8366403579711914, "num_tokens": 17998008.0, "step": 473 }, { "epoch": 0.06029767205190179, "ewc_loss": 0.00193023681640625, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 1.55717134475708e-06, "grad_norm": 5.106101036071777, "learning_rate": 2.0050869012293344e-07, "loss": 0.5178, "mean_token_accuracy": 0.8339268565177917, "num_tokens": 18032427.0, "step": 474 }, { "epoch": 0.060424882330492306, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 5.039268493652344, "learning_rate": 2.009325985587113e-07, "loss": 0.5487, "mean_token_accuracy": 0.8270992040634155, "num_tokens": 18069848.0, "step": 475 }, { "epoch": 0.060552092609082814, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 1.5944242477416992e-06, "grad_norm": 5.258268356323242, "learning_rate": 2.0135650699448918e-07, "loss": 0.6414, "mean_token_accuracy": 0.8012271523475647, "num_tokens": 18109240.0, "step": 476 }, { "epoch": 0.06067930288767332, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 5.0912041664123535, "learning_rate": 2.0178041543026706e-07, "loss": 0.6116, "mean_token_accuracy": 0.8120075464248657, "num_tokens": 18146849.0, "step": 477 }, { "epoch": 0.060806513166263836, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.7103891372680664e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 5.120375633239746, "learning_rate": 2.022043238660449e-07, "loss": 0.6417, "mean_token_accuracy": 0.7998157143592834, "num_tokens": 18186872.0, "step": 478 }, { "epoch": 0.06093372344485434, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.7401914596557617e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 4.822159767150879, "learning_rate": 2.026282323018228e-07, "loss": 0.6174, "mean_token_accuracy": 0.8101038336753845, "num_tokens": 18222299.0, "step": 479 }, { "epoch": 0.06106093372344486, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.7401914596557617e-06, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 4.529253005981445, "learning_rate": 2.0305214073760065e-07, "loss": 0.6211, "mean_token_accuracy": 0.8109316825866699, "num_tokens": 18261645.0, "step": 480 }, { "epoch": 0.061188144002035365, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.769993782043457e-06, "ewc_loss_parallel": 1.564621925354004e-06, "grad_norm": 4.959184646606445, "learning_rate": 2.0347604917337855e-07, "loss": 0.5919, "mean_token_accuracy": 0.8162139654159546, "num_tokens": 18292970.0, "step": 481 }, { "epoch": 0.06131535428062587, "ewc_loss": 0.0019989013671875, "ewc_loss_diag": 3.7550926208496094e-06, "ewc_loss_parallel": 1.6242265701293945e-06, "grad_norm": 4.696336269378662, "learning_rate": 2.038999576091564e-07, "loss": 0.5391, "mean_token_accuracy": 0.8336536884307861, "num_tokens": 18331133.0, "step": 482 }, { "epoch": 0.06144256455921639, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.7997961044311523e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 4.875480651855469, "learning_rate": 2.043238660449343e-07, "loss": 0.6254, "mean_token_accuracy": 0.8065420985221863, "num_tokens": 18368905.0, "step": 483 }, { "epoch": 0.061569774837806894, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.6242265701293945e-06, "grad_norm": 4.845678329467773, "learning_rate": 2.0474777448071214e-07, "loss": 0.5502, "mean_token_accuracy": 0.8296281695365906, "num_tokens": 18406362.0, "step": 484 }, { "epoch": 0.0616969851163974, "ewc_loss": 0.0019989013671875, "ewc_loss_diag": 3.7848949432373047e-06, "ewc_loss_parallel": 1.6242265701293945e-06, "grad_norm": 4.984676361083984, "learning_rate": 2.0517168291649004e-07, "loss": 0.5871, "mean_token_accuracy": 0.8183193802833557, "num_tokens": 18439246.0, "step": 485 }, { "epoch": 0.061824195394987916, "ewc_loss": 0.002044677734375, "ewc_loss_diag": 3.7848949432373047e-06, "ewc_loss_parallel": 1.6614794731140137e-06, "grad_norm": 5.536805152893066, "learning_rate": 2.0559559135226788e-07, "loss": 0.5911, "mean_token_accuracy": 0.8139539361000061, "num_tokens": 18473408.0, "step": 486 }, { "epoch": 0.06195140567357842, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.7136335372924805e-06, "grad_norm": 5.025711536407471, "learning_rate": 2.0601949978804578e-07, "loss": 0.5448, "mean_token_accuracy": 0.8282844424247742, "num_tokens": 18513246.0, "step": 487 }, { "epoch": 0.06207861595216894, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.7997961044311523e-06, "ewc_loss_parallel": 1.6540288925170898e-06, "grad_norm": 5.060548782348633, "learning_rate": 2.0644340822382363e-07, "loss": 0.5171, "mean_token_accuracy": 0.8361997604370117, "num_tokens": 18549125.0, "step": 488 }, { "epoch": 0.062205826230759445, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.646578311920166e-06, "grad_norm": 5.1239447593688965, "learning_rate": 2.0686731665960153e-07, "loss": 0.5672, "mean_token_accuracy": 0.8279436826705933, "num_tokens": 18585306.0, "step": 489 }, { "epoch": 0.06233303650934995, "ewc_loss": 0.002044677734375, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6540288925170898e-06, "grad_norm": 5.186706066131592, "learning_rate": 2.0729122509537937e-07, "loss": 0.5266, "mean_token_accuracy": 0.835918664932251, "num_tokens": 18624378.0, "step": 490 }, { "epoch": 0.06246024678794047, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.691281795501709e-06, "grad_norm": 5.426666259765625, "learning_rate": 2.0771513353115727e-07, "loss": 0.5702, "mean_token_accuracy": 0.821835994720459, "num_tokens": 18660814.0, "step": 491 }, { "epoch": 0.06258745706653097, "ewc_loss": 0.0020904541015625, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.7061829566955566e-06, "grad_norm": 4.6900458335876465, "learning_rate": 2.0813904196693512e-07, "loss": 0.5385, "mean_token_accuracy": 0.8321585655212402, "num_tokens": 18702129.0, "step": 492 }, { "epoch": 0.06271466734512149, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6316771507263184e-06, "grad_norm": 5.3548784255981445, "learning_rate": 2.0856295040271302e-07, "loss": 0.6099, "mean_token_accuracy": 0.8087255954742432, "num_tokens": 18737496.0, "step": 493 }, { "epoch": 0.06284187762371199, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7136335372924805e-06, "grad_norm": 4.83798885345459, "learning_rate": 2.0898685883849086e-07, "loss": 0.5673, "mean_token_accuracy": 0.8258320093154907, "num_tokens": 18779332.0, "step": 494 }, { "epoch": 0.0629690879023025, "ewc_loss": 0.002044677734375, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 1.6614794731140137e-06, "grad_norm": 4.704219341278076, "learning_rate": 2.0941076727426874e-07, "loss": 0.5529, "mean_token_accuracy": 0.8266599178314209, "num_tokens": 18817976.0, "step": 495 }, { "epoch": 0.06309629818089302, "ewc_loss": 0.002044677734375, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6540288925170898e-06, "grad_norm": 4.712136268615723, "learning_rate": 2.098346757100466e-07, "loss": 0.5437, "mean_token_accuracy": 0.8288601040840149, "num_tokens": 18862860.0, "step": 496 }, { "epoch": 0.06322350845948353, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.6838312149047852e-06, "grad_norm": 4.8958420753479, "learning_rate": 2.1025858414582448e-07, "loss": 0.5844, "mean_token_accuracy": 0.816857099533081, "num_tokens": 18899025.0, "step": 497 }, { "epoch": 0.06335071873807403, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7061829566955566e-06, "grad_norm": 4.971155643463135, "learning_rate": 2.1068249258160238e-07, "loss": 0.5506, "mean_token_accuracy": 0.8209781646728516, "num_tokens": 18933118.0, "step": 498 }, { "epoch": 0.06347792901666455, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7136335372924805e-06, "grad_norm": 5.214478969573975, "learning_rate": 2.1110640101738023e-07, "loss": 0.5513, "mean_token_accuracy": 0.8288775682449341, "num_tokens": 18969165.0, "step": 499 }, { "epoch": 0.06360513929525506, "ewc_loss": 0.0021209716796875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7285346984863281e-06, "grad_norm": 5.094918251037598, "learning_rate": 2.1153030945315813e-07, "loss": 0.5607, "mean_token_accuracy": 0.8238393664360046, "num_tokens": 19003882.0, "step": 500 }, { "epoch": 0.06373234957384556, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7434358596801758e-06, "grad_norm": 5.302716255187988, "learning_rate": 2.1195421788893597e-07, "loss": 0.561, "mean_token_accuracy": 0.8254420757293701, "num_tokens": 19037540.0, "step": 501 }, { "epoch": 0.06385955985243608, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7508864402770996e-06, "grad_norm": 4.912245273590088, "learning_rate": 2.1237812632471387e-07, "loss": 0.5712, "mean_token_accuracy": 0.8168437480926514, "num_tokens": 19077270.0, "step": 502 }, { "epoch": 0.06398677013102659, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7061829566955566e-06, "grad_norm": 4.809145927429199, "learning_rate": 2.1280203476049172e-07, "loss": 0.6589, "mean_token_accuracy": 0.792078971862793, "num_tokens": 19117579.0, "step": 503 }, { "epoch": 0.06411398040961709, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7061829566955566e-06, "grad_norm": 4.808787822723389, "learning_rate": 2.1322594319626962e-07, "loss": 0.5988, "mean_token_accuracy": 0.813017725944519, "num_tokens": 19156599.0, "step": 504 }, { "epoch": 0.0642411906882076, "ewc_loss": 0.0021209716796875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7210841178894043e-06, "grad_norm": 5.356914520263672, "learning_rate": 2.1364985163204746e-07, "loss": 0.5898, "mean_token_accuracy": 0.8157346248626709, "num_tokens": 19187367.0, "step": 505 }, { "epoch": 0.06436840096679812, "ewc_loss": 0.002197265625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.8030405044555664e-06, "grad_norm": 5.064190864562988, "learning_rate": 2.1407376006782536e-07, "loss": 0.5245, "mean_token_accuracy": 0.8343135118484497, "num_tokens": 19223916.0, "step": 506 }, { "epoch": 0.06449561124538863, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7434358596801758e-06, "grad_norm": 4.970034599304199, "learning_rate": 2.144976685036032e-07, "loss": 0.5395, "mean_token_accuracy": 0.8312124013900757, "num_tokens": 19260336.0, "step": 507 }, { "epoch": 0.06462282152397913, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 5.014003276824951, "learning_rate": 2.149215769393811e-07, "loss": 0.6022, "mean_token_accuracy": 0.8143786191940308, "num_tokens": 19299895.0, "step": 508 }, { "epoch": 0.06475003180256965, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7583370208740234e-06, "grad_norm": 4.986868381500244, "learning_rate": 2.1534548537515895e-07, "loss": 0.5432, "mean_token_accuracy": 0.8287374973297119, "num_tokens": 19336614.0, "step": 509 }, { "epoch": 0.06487724208116016, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 4.829402446746826, "learning_rate": 2.1576939381093685e-07, "loss": 0.5941, "mean_token_accuracy": 0.8164457678794861, "num_tokens": 19378683.0, "step": 510 }, { "epoch": 0.06500445235975066, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7136335372924805e-06, "grad_norm": 4.812399864196777, "learning_rate": 2.161933022467147e-07, "loss": 0.578, "mean_token_accuracy": 0.8164783716201782, "num_tokens": 19422598.0, "step": 511 }, { "epoch": 0.06513166263834118, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7657876014709473e-06, "grad_norm": 4.66743278503418, "learning_rate": 2.166172106824926e-07, "loss": 0.5683, "mean_token_accuracy": 0.8253052234649658, "num_tokens": 19466972.0, "step": 512 }, { "epoch": 0.06525887291693169, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 5.202920436859131, "learning_rate": 2.1704111911827044e-07, "loss": 0.559, "mean_token_accuracy": 0.8199799656867981, "num_tokens": 19503809.0, "step": 513 }, { "epoch": 0.0653860831955222, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.817941665649414e-06, "grad_norm": 4.962306976318359, "learning_rate": 2.1746502755404831e-07, "loss": 0.6145, "mean_token_accuracy": 0.814720869064331, "num_tokens": 19545713.0, "step": 514 }, { "epoch": 0.06551329347411271, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7955899238586426e-06, "grad_norm": 4.866228103637695, "learning_rate": 2.178889359898262e-07, "loss": 0.6008, "mean_token_accuracy": 0.8161488771438599, "num_tokens": 19582272.0, "step": 515 }, { "epoch": 0.06564050375270322, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7434358596801758e-06, "grad_norm": 4.8309478759765625, "learning_rate": 2.1831284442560406e-07, "loss": 0.6166, "mean_token_accuracy": 0.8101223111152649, "num_tokens": 19625574.0, "step": 516 }, { "epoch": 0.06576771403129372, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.7583370208740234e-06, "grad_norm": 5.139536380767822, "learning_rate": 2.1873675286138193e-07, "loss": 0.6038, "mean_token_accuracy": 0.8098797798156738, "num_tokens": 19667791.0, "step": 517 }, { "epoch": 0.06589492430988424, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.817941665649414e-06, "grad_norm": 4.884011745452881, "learning_rate": 2.191606612971598e-07, "loss": 0.6324, "mean_token_accuracy": 0.8049004077911377, "num_tokens": 19708289.0, "step": 518 }, { "epoch": 0.06602213458847475, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.7881393432617188e-06, "grad_norm": 4.717060089111328, "learning_rate": 2.1958456973293768e-07, "loss": 0.6114, "mean_token_accuracy": 0.805600106716156, "num_tokens": 19747444.0, "step": 519 }, { "epoch": 0.06614934486706527, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.7583370208740234e-06, "grad_norm": 4.871138572692871, "learning_rate": 2.2000847816871555e-07, "loss": 0.4996, "mean_token_accuracy": 0.8379684686660767, "num_tokens": 19787061.0, "step": 520 }, { "epoch": 0.06627655514565577, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.7955899238586426e-06, "grad_norm": 4.875734806060791, "learning_rate": 2.2043238660449342e-07, "loss": 0.5417, "mean_token_accuracy": 0.8313884735107422, "num_tokens": 19826893.0, "step": 521 }, { "epoch": 0.06640376542424628, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.7955899238586426e-06, "grad_norm": 4.948230743408203, "learning_rate": 2.208562950402713e-07, "loss": 0.5348, "mean_token_accuracy": 0.8286239504814148, "num_tokens": 19864285.0, "step": 522 }, { "epoch": 0.0665309757028368, "ewc_loss": 0.002227783203125, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 1.8328428268432617e-06, "grad_norm": 4.725166320800781, "learning_rate": 2.2128020347604917e-07, "loss": 0.5699, "mean_token_accuracy": 0.8221023082733154, "num_tokens": 19903806.0, "step": 523 }, { "epoch": 0.0666581859814273, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.780688762664795e-06, "grad_norm": 5.080228805541992, "learning_rate": 2.2170411191182704e-07, "loss": 0.5477, "mean_token_accuracy": 0.8263019323348999, "num_tokens": 19950382.0, "step": 524 }, { "epoch": 0.06678539626001781, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8477439880371094e-06, "grad_norm": 4.893163204193115, "learning_rate": 2.221280203476049e-07, "loss": 0.5713, "mean_token_accuracy": 0.8184877634048462, "num_tokens": 19984765.0, "step": 525 }, { "epoch": 0.06691260653860832, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8104910850524902e-06, "grad_norm": 4.729238033294678, "learning_rate": 2.2255192878338279e-07, "loss": 0.6113, "mean_token_accuracy": 0.812043309211731, "num_tokens": 20021081.0, "step": 526 }, { "epoch": 0.06703981681719882, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.7881393432617188e-06, "grad_norm": 5.386192798614502, "learning_rate": 2.2297583721916066e-07, "loss": 0.5841, "mean_token_accuracy": 0.8169929385185242, "num_tokens": 20055860.0, "step": 527 }, { "epoch": 0.06716702709578934, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8849968910217285e-06, "grad_norm": 5.048342227935791, "learning_rate": 2.2339974565493853e-07, "loss": 0.5392, "mean_token_accuracy": 0.8317059278488159, "num_tokens": 20095023.0, "step": 528 }, { "epoch": 0.06729423737437985, "ewc_loss": 0.002227783203125, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8253922462463379e-06, "grad_norm": 4.868789196014404, "learning_rate": 2.238236540907164e-07, "loss": 0.5932, "mean_token_accuracy": 0.8155468702316284, "num_tokens": 20134620.0, "step": 529 }, { "epoch": 0.06742144765297035, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 4.023313522338867e-06, "ewc_loss_parallel": 1.8104910850524902e-06, "grad_norm": 5.154356956481934, "learning_rate": 2.2424756252649428e-07, "loss": 0.5643, "mean_token_accuracy": 0.8212342262268066, "num_tokens": 20169690.0, "step": 530 }, { "epoch": 0.06754865793156087, "ewc_loss": 0.00225830078125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.8477439880371094e-06, "grad_norm": 5.238564968109131, "learning_rate": 2.2467147096227215e-07, "loss": 0.5564, "mean_token_accuracy": 0.824964165687561, "num_tokens": 20203609.0, "step": 531 }, { "epoch": 0.06767586821015138, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.8700957298278809e-06, "grad_norm": 4.935090065002441, "learning_rate": 2.2509537939805002e-07, "loss": 0.563, "mean_token_accuracy": 0.8218706846237183, "num_tokens": 20234616.0, "step": 532 }, { "epoch": 0.0678030784887419, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 1.8328428268432617e-06, "grad_norm": 4.715332508087158, "learning_rate": 2.2551928783382787e-07, "loss": 0.5528, "mean_token_accuracy": 0.82352215051651, "num_tokens": 20271634.0, "step": 533 }, { "epoch": 0.0679302887673324, "ewc_loss": 0.002227783203125, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 1.817941665649414e-06, "grad_norm": 5.011786460876465, "learning_rate": 2.2594319626960577e-07, "loss": 0.5808, "mean_token_accuracy": 0.8224530816078186, "num_tokens": 20309683.0, "step": 534 }, { "epoch": 0.06805749904592291, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 4.796253204345703, "learning_rate": 2.263671047053836e-07, "loss": 0.4956, "mean_token_accuracy": 0.8380610942840576, "num_tokens": 20346013.0, "step": 535 }, { "epoch": 0.06818470932451343, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 1.8328428268432617e-06, "grad_norm": 4.978104114532471, "learning_rate": 2.267910131411615e-07, "loss": 0.5526, "mean_token_accuracy": 0.8287243843078613, "num_tokens": 20387613.0, "step": 536 }, { "epoch": 0.06831191960310393, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 1.8849968910217285e-06, "grad_norm": 5.090679168701172, "learning_rate": 2.2721492157693936e-07, "loss": 0.5515, "mean_token_accuracy": 0.82636559009552, "num_tokens": 20417579.0, "step": 537 }, { "epoch": 0.06843912988169444, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 4.808280944824219, "learning_rate": 2.2763883001271726e-07, "loss": 0.5243, "mean_token_accuracy": 0.8328021764755249, "num_tokens": 20455100.0, "step": 538 }, { "epoch": 0.06856634016028496, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.8551945686340332e-06, "grad_norm": 4.780041694641113, "learning_rate": 2.280627384484951e-07, "loss": 0.5861, "mean_token_accuracy": 0.8141937255859375, "num_tokens": 20496458.0, "step": 539 }, { "epoch": 0.06869355043887546, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 1.862645149230957e-06, "grad_norm": 5.252395153045654, "learning_rate": 2.28486646884273e-07, "loss": 0.5607, "mean_token_accuracy": 0.822527289390564, "num_tokens": 20533758.0, "step": 540 }, { "epoch": 0.06882076071746597, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.9371509552001953e-06, "grad_norm": 5.127631187438965, "learning_rate": 2.2891055532005085e-07, "loss": 0.6056, "mean_token_accuracy": 0.8132216930389404, "num_tokens": 20567259.0, "step": 541 }, { "epoch": 0.06894797099605648, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.9073486328125e-06, "grad_norm": 5.133263111114502, "learning_rate": 2.2933446375582875e-07, "loss": 0.5865, "mean_token_accuracy": 0.8189108371734619, "num_tokens": 20599333.0, "step": 542 }, { "epoch": 0.06907518127464699, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 1.9073486328125e-06, "grad_norm": 4.98649263381958, "learning_rate": 2.297583721916066e-07, "loss": 0.5236, "mean_token_accuracy": 0.8361847400665283, "num_tokens": 20640195.0, "step": 543 }, { "epoch": 0.0692023915532375, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 1.9073486328125e-06, "grad_norm": 4.902453899383545, "learning_rate": 2.301822806273845e-07, "loss": 0.5441, "mean_token_accuracy": 0.8286699056625366, "num_tokens": 20682617.0, "step": 544 }, { "epoch": 0.06932960183182801, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.8998980522155762e-06, "grad_norm": 5.271191120147705, "learning_rate": 2.3060618906316234e-07, "loss": 0.5962, "mean_token_accuracy": 0.8161634206771851, "num_tokens": 20718032.0, "step": 545 }, { "epoch": 0.06945681211041853, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 1.9371509552001953e-06, "grad_norm": 5.005950450897217, "learning_rate": 2.3103009749894024e-07, "loss": 0.5606, "mean_token_accuracy": 0.824293315410614, "num_tokens": 20755190.0, "step": 546 }, { "epoch": 0.06958402238900903, "ewc_loss": 0.0023345947265625, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.9222497940063477e-06, "grad_norm": 4.710731506347656, "learning_rate": 2.3145400593471808e-07, "loss": 0.5853, "mean_token_accuracy": 0.816250205039978, "num_tokens": 20799088.0, "step": 547 }, { "epoch": 0.06971123266759954, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.8998980522155762e-06, "grad_norm": 5.339086532592773, "learning_rate": 2.3187791437049598e-07, "loss": 0.6158, "mean_token_accuracy": 0.8059713840484619, "num_tokens": 20831819.0, "step": 548 }, { "epoch": 0.06983844294619006, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.3737406730651855, "learning_rate": 2.3230182280627383e-07, "loss": 0.5456, "mean_token_accuracy": 0.829552173614502, "num_tokens": 20861291.0, "step": 549 }, { "epoch": 0.06996565322478056, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.757780075073242, "learning_rate": 2.327257312420517e-07, "loss": 0.5224, "mean_token_accuracy": 0.8369432687759399, "num_tokens": 20897446.0, "step": 550 }, { "epoch": 0.07009286350337107, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.8998980522155762e-06, "grad_norm": 5.389503002166748, "learning_rate": 2.3314963967782957e-07, "loss": 0.5456, "mean_token_accuracy": 0.8243451118469238, "num_tokens": 20934423.0, "step": 551 }, { "epoch": 0.07022007378196159, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.160529613494873, "learning_rate": 2.3357354811360745e-07, "loss": 0.5636, "mean_token_accuracy": 0.826141893863678, "num_tokens": 20978394.0, "step": 552 }, { "epoch": 0.07034728406055209, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.285330295562744, "learning_rate": 2.3399745654938532e-07, "loss": 0.5896, "mean_token_accuracy": 0.8166904449462891, "num_tokens": 21010669.0, "step": 553 }, { "epoch": 0.0704744943391426, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.291534423828125e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 4.845188617706299, "learning_rate": 2.344213649851632e-07, "loss": 0.5587, "mean_token_accuracy": 0.8261900544166565, "num_tokens": 21047696.0, "step": 554 }, { "epoch": 0.07060170461773312, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 4.291534423828125e-06, "ewc_loss_parallel": 1.9222497940063477e-06, "grad_norm": 4.756583213806152, "learning_rate": 2.3484527342094106e-07, "loss": 0.6397, "mean_token_accuracy": 0.8042043447494507, "num_tokens": 21091437.0, "step": 555 }, { "epoch": 0.07072891489632362, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 1.9222497940063477e-06, "grad_norm": 4.881890773773193, "learning_rate": 2.3526918185671894e-07, "loss": 0.6007, "mean_token_accuracy": 0.8131278157234192, "num_tokens": 21128880.0, "step": 556 }, { "epoch": 0.07085612517491413, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.291534423828125e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 4.9518818855285645, "learning_rate": 2.356930902924968e-07, "loss": 0.5524, "mean_token_accuracy": 0.8277359008789062, "num_tokens": 21168523.0, "step": 557 }, { "epoch": 0.07098333545350465, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.85757303237915, "learning_rate": 2.3611699872827468e-07, "loss": 0.6383, "mean_token_accuracy": 0.8010154366493225, "num_tokens": 21211376.0, "step": 558 }, { "epoch": 0.07111054573209516, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9371509552001953e-06, "grad_norm": 4.798537731170654, "learning_rate": 2.3654090716405255e-07, "loss": 0.5709, "mean_token_accuracy": 0.8211564421653748, "num_tokens": 21249162.0, "step": 559 }, { "epoch": 0.07123775601068566, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 5.1345062255859375, "learning_rate": 2.3696481559983043e-07, "loss": 0.5977, "mean_token_accuracy": 0.8128116726875305, "num_tokens": 21282351.0, "step": 560 }, { "epoch": 0.07136496628927617, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.864477634429932, "learning_rate": 2.373887240356083e-07, "loss": 0.5882, "mean_token_accuracy": 0.8167881965637207, "num_tokens": 21319508.0, "step": 561 }, { "epoch": 0.07149217656786669, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.804800033569336, "learning_rate": 2.3781263247138617e-07, "loss": 0.5529, "mean_token_accuracy": 0.8278682231903076, "num_tokens": 21357111.0, "step": 562 }, { "epoch": 0.07161938684645719, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 5.007776737213135, "learning_rate": 2.3823654090716404e-07, "loss": 0.5546, "mean_token_accuracy": 0.8231289386749268, "num_tokens": 21393222.0, "step": 563 }, { "epoch": 0.0717465971250477, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.801960468292236, "learning_rate": 2.386604493429419e-07, "loss": 0.5095, "mean_token_accuracy": 0.8370039463043213, "num_tokens": 21429595.0, "step": 564 }, { "epoch": 0.07187380740363822, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.062302112579346, "learning_rate": 2.390843577787198e-07, "loss": 0.6052, "mean_token_accuracy": 0.8066507577896118, "num_tokens": 21467341.0, "step": 565 }, { "epoch": 0.07200101768222872, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.763572692871094, "learning_rate": 2.3950826621449766e-07, "loss": 0.4832, "mean_token_accuracy": 0.847730278968811, "num_tokens": 21510927.0, "step": 566 }, { "epoch": 0.07212822796081923, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.893629550933838, "learning_rate": 2.3993217465027556e-07, "loss": 0.5587, "mean_token_accuracy": 0.8257864713668823, "num_tokens": 21551604.0, "step": 567 }, { "epoch": 0.07225543823940975, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 5.227968692779541, "learning_rate": 2.403560830860534e-07, "loss": 0.609, "mean_token_accuracy": 0.8135198950767517, "num_tokens": 21591871.0, "step": 568 }, { "epoch": 0.07238264851800025, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 5.072107791900635, "learning_rate": 2.4077999152183125e-07, "loss": 0.6065, "mean_token_accuracy": 0.8106215000152588, "num_tokens": 21629810.0, "step": 569 }, { "epoch": 0.07250985879659076, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.982158660888672, "learning_rate": 2.4120389995760915e-07, "loss": 0.5053, "mean_token_accuracy": 0.8430810570716858, "num_tokens": 21668706.0, "step": 570 }, { "epoch": 0.07263706907518128, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 5.150697231292725, "learning_rate": 2.41627808393387e-07, "loss": 0.5494, "mean_token_accuracy": 0.8250138163566589, "num_tokens": 21712725.0, "step": 571 }, { "epoch": 0.07276427935377179, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.95330810546875, "learning_rate": 2.420517168291649e-07, "loss": 0.5597, "mean_token_accuracy": 0.8266419172286987, "num_tokens": 21746086.0, "step": 572 }, { "epoch": 0.07289148963236229, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.960714817047119, "learning_rate": 2.4247562526494274e-07, "loss": 0.5195, "mean_token_accuracy": 0.8337888717651367, "num_tokens": 21779573.0, "step": 573 }, { "epoch": 0.0730186999109528, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.849174499511719, "learning_rate": 2.4289953370072064e-07, "loss": 0.5913, "mean_token_accuracy": 0.8140692710876465, "num_tokens": 21818766.0, "step": 574 }, { "epoch": 0.07314591018954332, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.804835796356201, "learning_rate": 2.433234421364985e-07, "loss": 0.5471, "mean_token_accuracy": 0.8315419554710388, "num_tokens": 21855539.0, "step": 575 }, { "epoch": 0.07327312046813382, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.794590473175049, "learning_rate": 2.437473505722764e-07, "loss": 0.5123, "mean_token_accuracy": 0.8354766368865967, "num_tokens": 21892567.0, "step": 576 }, { "epoch": 0.07340033074672433, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.917277812957764, "learning_rate": 2.4417125900805423e-07, "loss": 0.5309, "mean_token_accuracy": 0.8248804211616516, "num_tokens": 21928607.0, "step": 577 }, { "epoch": 0.07352754102531485, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 5.047853946685791, "learning_rate": 2.4459516744383213e-07, "loss": 0.5781, "mean_token_accuracy": 0.8194167017936707, "num_tokens": 21963781.0, "step": 578 }, { "epoch": 0.07365475130390535, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.9326558113098145, "learning_rate": 2.4501907587961e-07, "loss": 0.5387, "mean_token_accuracy": 0.8309193849563599, "num_tokens": 21999140.0, "step": 579 }, { "epoch": 0.07378196158249586, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.77655553817749, "learning_rate": 2.454429843153879e-07, "loss": 0.6377, "mean_token_accuracy": 0.8005045652389526, "num_tokens": 22038703.0, "step": 580 }, { "epoch": 0.07390917186108638, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 5.111605644226074, "learning_rate": 2.458668927511657e-07, "loss": 0.5523, "mean_token_accuracy": 0.8196408748626709, "num_tokens": 22071052.0, "step": 581 }, { "epoch": 0.07403638213967688, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.88031005859375, "learning_rate": 2.462908011869436e-07, "loss": 0.5172, "mean_token_accuracy": 0.8365795016288757, "num_tokens": 22109629.0, "step": 582 }, { "epoch": 0.0741635924182674, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.5299530029296875e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.800776481628418, "learning_rate": 2.4671470962272147e-07, "loss": 0.5297, "mean_token_accuracy": 0.8335587978363037, "num_tokens": 22151544.0, "step": 583 }, { "epoch": 0.07429080269685791, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.5299530029296875e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 5.021677494049072, "learning_rate": 2.4713861805849937e-07, "loss": 0.5075, "mean_token_accuracy": 0.8362764120101929, "num_tokens": 22189241.0, "step": 584 }, { "epoch": 0.07441801297544842, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.885261535644531, "learning_rate": 2.475625264942772e-07, "loss": 0.523, "mean_token_accuracy": 0.8338955640792847, "num_tokens": 22229138.0, "step": 585 }, { "epoch": 0.07454522325403892, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.5299530029296875e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.9327239990234375, "learning_rate": 2.479864349300551e-07, "loss": 0.5613, "mean_token_accuracy": 0.8249623775482178, "num_tokens": 22264930.0, "step": 586 }, { "epoch": 0.07467243353262944, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 4.5299530029296875e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 5.0005974769592285, "learning_rate": 2.4841034336583296e-07, "loss": 0.5035, "mean_token_accuracy": 0.8347153067588806, "num_tokens": 22297162.0, "step": 587 }, { "epoch": 0.07479964381121995, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 4.80863618850708, "learning_rate": 2.488342518016108e-07, "loss": 0.5865, "mean_token_accuracy": 0.813152551651001, "num_tokens": 22334779.0, "step": 588 }, { "epoch": 0.07492685408981045, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 4.973785877227783, "learning_rate": 2.492581602373887e-07, "loss": 0.5438, "mean_token_accuracy": 0.830045223236084, "num_tokens": 22372820.0, "step": 589 }, { "epoch": 0.07505406436840097, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 5.033195972442627, "learning_rate": 2.4968206867316655e-07, "loss": 0.5642, "mean_token_accuracy": 0.8218438625335693, "num_tokens": 22417841.0, "step": 590 }, { "epoch": 0.07518127464699148, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 4.988980770111084, "learning_rate": 2.5010597710894445e-07, "loss": 0.479, "mean_token_accuracy": 0.8476570248603821, "num_tokens": 22449133.0, "step": 591 }, { "epoch": 0.07530848492558198, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.849426746368408, "learning_rate": 2.505298855447223e-07, "loss": 0.5438, "mean_token_accuracy": 0.8291490077972412, "num_tokens": 22491145.0, "step": 592 }, { "epoch": 0.0754356952041725, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.6193599700927734e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.867568016052246, "learning_rate": 2.509537939805002e-07, "loss": 0.5568, "mean_token_accuracy": 0.8271539211273193, "num_tokens": 22531738.0, "step": 593 }, { "epoch": 0.07556290548276301, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.6193599700927734e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 5.132882118225098, "learning_rate": 2.513777024162781e-07, "loss": 0.5355, "mean_token_accuracy": 0.8329600691795349, "num_tokens": 22563005.0, "step": 594 }, { "epoch": 0.07569011576135352, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 4.649162292480469e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.804512977600098, "learning_rate": 2.5180161085205594e-07, "loss": 0.5172, "mean_token_accuracy": 0.8377971649169922, "num_tokens": 22601333.0, "step": 595 }, { "epoch": 0.07581732603994402, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.899866580963135, "learning_rate": 2.522255192878338e-07, "loss": 0.5705, "mean_token_accuracy": 0.8212432861328125, "num_tokens": 22639528.0, "step": 596 }, { "epoch": 0.07594453631853454, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 5.00829553604126, "learning_rate": 2.526494277236117e-07, "loss": 0.5538, "mean_token_accuracy": 0.8254978656768799, "num_tokens": 22675290.0, "step": 597 }, { "epoch": 0.07607174659712505, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 5.093142986297607, "learning_rate": 2.530733361593896e-07, "loss": 0.5291, "mean_token_accuracy": 0.8311991691589355, "num_tokens": 22712823.0, "step": 598 }, { "epoch": 0.07619895687571555, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.953547954559326, "learning_rate": 2.5349724459516743e-07, "loss": 0.5504, "mean_token_accuracy": 0.8280078172683716, "num_tokens": 22750916.0, "step": 599 }, { "epoch": 0.07632616715430607, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.971909046173096, "learning_rate": 2.539211530309453e-07, "loss": 0.5019, "mean_token_accuracy": 0.8385791182518005, "num_tokens": 22782537.0, "step": 600 }, { "epoch": 0.07645337743289658, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 5.112308979034424, "learning_rate": 2.543450614667232e-07, "loss": 0.5932, "mean_token_accuracy": 0.8158590793609619, "num_tokens": 22814805.0, "step": 601 }, { "epoch": 0.07658058771148708, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.937435150146484, "learning_rate": 2.547689699025011e-07, "loss": 0.5917, "mean_token_accuracy": 0.8138842582702637, "num_tokens": 22852654.0, "step": 602 }, { "epoch": 0.0767077979900776, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.898483753204346, "learning_rate": 2.551928783382789e-07, "loss": 0.5459, "mean_token_accuracy": 0.8255727887153625, "num_tokens": 22894205.0, "step": 603 }, { "epoch": 0.07683500826866811, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.916652679443359, "learning_rate": 2.5561678677405677e-07, "loss": 0.5124, "mean_token_accuracy": 0.835356593132019, "num_tokens": 22924095.0, "step": 604 }, { "epoch": 0.07696221854725861, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 4.932967185974121, "learning_rate": 2.5604069520983467e-07, "loss": 0.5945, "mean_token_accuracy": 0.8109567761421204, "num_tokens": 22960857.0, "step": 605 }, { "epoch": 0.07708942882584913, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.803966522216797, "learning_rate": 2.564646036456125e-07, "loss": 0.547, "mean_token_accuracy": 0.8251507878303528, "num_tokens": 23004670.0, "step": 606 }, { "epoch": 0.07721663910443964, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 5.02387809753418, "learning_rate": 2.568885120813904e-07, "loss": 0.5466, "mean_token_accuracy": 0.8250557780265808, "num_tokens": 23041999.0, "step": 607 }, { "epoch": 0.07734384938303016, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.919068813323975, "learning_rate": 2.5731242051716826e-07, "loss": 0.5354, "mean_token_accuracy": 0.8330617547035217, "num_tokens": 23080704.0, "step": 608 }, { "epoch": 0.07747105966162066, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.958871364593506, "learning_rate": 2.5773632895294616e-07, "loss": 0.5705, "mean_token_accuracy": 0.8220529556274414, "num_tokens": 23118662.0, "step": 609 }, { "epoch": 0.07759826994021117, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.945526123046875, "learning_rate": 2.58160237388724e-07, "loss": 0.4935, "mean_token_accuracy": 0.8392874002456665, "num_tokens": 23150367.0, "step": 610 }, { "epoch": 0.07772548021880168, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.997418403625488, "learning_rate": 2.585841458245019e-07, "loss": 0.5514, "mean_token_accuracy": 0.8255221247673035, "num_tokens": 23185836.0, "step": 611 }, { "epoch": 0.07785269049739219, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.882867813110352, "learning_rate": 2.5900805426027975e-07, "loss": 0.4926, "mean_token_accuracy": 0.8416181802749634, "num_tokens": 23219346.0, "step": 612 }, { "epoch": 0.0779799007759827, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 4.96855354309082, "learning_rate": 2.5943196269605765e-07, "loss": 0.5639, "mean_token_accuracy": 0.8274404406547546, "num_tokens": 23257569.0, "step": 613 }, { "epoch": 0.07810711105457321, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.798173904418945e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.775395393371582, "learning_rate": 2.598558711318355e-07, "loss": 0.5542, "mean_token_accuracy": 0.8279785513877869, "num_tokens": 23302944.0, "step": 614 }, { "epoch": 0.07823432133316371, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.868224143981934, "learning_rate": 2.602797795676134e-07, "loss": 0.5602, "mean_token_accuracy": 0.8237006068229675, "num_tokens": 23341660.0, "step": 615 }, { "epoch": 0.07836153161175423, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 5.132782459259033, "learning_rate": 2.6070368800339124e-07, "loss": 0.5576, "mean_token_accuracy": 0.8190604448318481, "num_tokens": 23375956.0, "step": 616 }, { "epoch": 0.07848874189034474, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.827976226806641e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.946613788604736, "learning_rate": 2.6112759643916914e-07, "loss": 0.5101, "mean_token_accuracy": 0.8365273475646973, "num_tokens": 23407568.0, "step": 617 }, { "epoch": 0.07861595216893524, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.89225959777832, "learning_rate": 2.61551504874947e-07, "loss": 0.5425, "mean_token_accuracy": 0.8294041156768799, "num_tokens": 23440261.0, "step": 618 }, { "epoch": 0.07874316244752576, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.798173904418945e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.9297871589660645, "learning_rate": 2.619754133107249e-07, "loss": 0.5247, "mean_token_accuracy": 0.8345956206321716, "num_tokens": 23481706.0, "step": 619 }, { "epoch": 0.07887037272611627, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.827976226806641e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 4.960860729217529, "learning_rate": 2.623993217465028e-07, "loss": 0.585, "mean_token_accuracy": 0.8124039173126221, "num_tokens": 23515410.0, "step": 620 }, { "epoch": 0.07899758300470679, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.824062824249268, "learning_rate": 2.6282323018228063e-07, "loss": 0.5691, "mean_token_accuracy": 0.8242082595825195, "num_tokens": 23554627.0, "step": 621 }, { "epoch": 0.07912479328329729, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.827976226806641e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.869912624359131, "learning_rate": 2.632471386180585e-07, "loss": 0.5921, "mean_token_accuracy": 0.8128769993782043, "num_tokens": 23593400.0, "step": 622 }, { "epoch": 0.0792520035618878, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 5.125310897827148, "learning_rate": 2.6367104705383637e-07, "loss": 0.5896, "mean_token_accuracy": 0.8123762607574463, "num_tokens": 23628242.0, "step": 623 }, { "epoch": 0.07937921384047832, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.927692413330078, "learning_rate": 2.6409495548961427e-07, "loss": 0.5759, "mean_token_accuracy": 0.8234261870384216, "num_tokens": 23671041.0, "step": 624 }, { "epoch": 0.07950642411906882, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 4.995822906494141, "learning_rate": 2.6451886392539206e-07, "loss": 0.5533, "mean_token_accuracy": 0.8280308246612549, "num_tokens": 23701904.0, "step": 625 }, { "epoch": 0.07963363439765933, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.915015697479248, "learning_rate": 2.6494277236116996e-07, "loss": 0.5175, "mean_token_accuracy": 0.8350173830986023, "num_tokens": 23742374.0, "step": 626 }, { "epoch": 0.07976084467624985, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 5.105804443359375, "learning_rate": 2.6536668079694786e-07, "loss": 0.623, "mean_token_accuracy": 0.8064348697662354, "num_tokens": 23780216.0, "step": 627 }, { "epoch": 0.07988805495484035, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.993163585662842, "learning_rate": 2.6579058923272576e-07, "loss": 0.4766, "mean_token_accuracy": 0.8498793840408325, "num_tokens": 23813641.0, "step": 628 }, { "epoch": 0.08001526523343086, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.868727684020996, "learning_rate": 2.6621449766850356e-07, "loss": 0.562, "mean_token_accuracy": 0.8236110210418701, "num_tokens": 23850189.0, "step": 629 }, { "epoch": 0.08014247551202137, "ewc_loss": 0.002685546875, "ewc_loss_diag": 4.947185516357422e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.871090888977051, "learning_rate": 2.6663840610428145e-07, "loss": 0.5165, "mean_token_accuracy": 0.8359792232513428, "num_tokens": 23884410.0, "step": 630 }, { "epoch": 0.08026968579061187, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 4.877511024475098, "learning_rate": 2.6706231454005935e-07, "loss": 0.5258, "mean_token_accuracy": 0.8340799808502197, "num_tokens": 23930139.0, "step": 631 }, { "epoch": 0.08039689606920239, "ewc_loss": 0.002685546875, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.794632911682129, "learning_rate": 2.6748622297583725e-07, "loss": 0.5243, "mean_token_accuracy": 0.8341002464294434, "num_tokens": 23973124.0, "step": 632 }, { "epoch": 0.0805241063477929, "ewc_loss": 0.002685546875, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.869568824768066, "learning_rate": 2.6791013141161505e-07, "loss": 0.5326, "mean_token_accuracy": 0.8335583209991455, "num_tokens": 24008676.0, "step": 633 }, { "epoch": 0.08065131662638342, "ewc_loss": 0.002685546875, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 5.074896812438965, "learning_rate": 2.6833403984739294e-07, "loss": 0.5314, "mean_token_accuracy": 0.8314041495323181, "num_tokens": 24044880.0, "step": 634 }, { "epoch": 0.08077852690497392, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.860992908477783, "learning_rate": 2.6875794828317084e-07, "loss": 0.5219, "mean_token_accuracy": 0.8382731676101685, "num_tokens": 24090140.0, "step": 635 }, { "epoch": 0.08090573718356443, "ewc_loss": 0.002685546875, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 5.018022060394287, "learning_rate": 2.6918185671894874e-07, "loss": 0.6062, "mean_token_accuracy": 0.8139028549194336, "num_tokens": 24123311.0, "step": 636 }, { "epoch": 0.08103294746215495, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.991069793701172, "learning_rate": 2.6960576515472654e-07, "loss": 0.6403, "mean_token_accuracy": 0.7993826866149902, "num_tokens": 24164019.0, "step": 637 }, { "epoch": 0.08116015774074545, "ewc_loss": 0.002685546875, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.9735331535339355, "learning_rate": 2.7002967359050443e-07, "loss": 0.5073, "mean_token_accuracy": 0.8393922448158264, "num_tokens": 24205986.0, "step": 638 }, { "epoch": 0.08128736801933596, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 5.066394805908203e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.9576826095581055, "learning_rate": 2.7045358202628233e-07, "loss": 0.5646, "mean_token_accuracy": 0.8190039396286011, "num_tokens": 24251956.0, "step": 639 }, { "epoch": 0.08141457829792648, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 5.066394805908203e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.939600467681885, "learning_rate": 2.7087749046206023e-07, "loss": 0.5107, "mean_token_accuracy": 0.8358415365219116, "num_tokens": 24284990.0, "step": 640 }, { "epoch": 0.08154178857651698, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 5.066394805908203e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.971881866455078, "learning_rate": 2.71301398897838e-07, "loss": 0.5662, "mean_token_accuracy": 0.820786714553833, "num_tokens": 24323305.0, "step": 641 }, { "epoch": 0.08166899885510749, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.066394805908203e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.845327377319336, "learning_rate": 2.717253073336159e-07, "loss": 0.4784, "mean_token_accuracy": 0.8464584350585938, "num_tokens": 24365847.0, "step": 642 }, { "epoch": 0.081796209133698, "ewc_loss": 0.0027008056640625, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.9748945236206055, "learning_rate": 2.721492157693938e-07, "loss": 0.485, "mean_token_accuracy": 0.8448724746704102, "num_tokens": 24407269.0, "step": 643 }, { "epoch": 0.0819234194122885, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 4.986354827880859, "learning_rate": 2.7257312420517167e-07, "loss": 0.5416, "mean_token_accuracy": 0.8274233937263489, "num_tokens": 24444990.0, "step": 644 }, { "epoch": 0.08205062969087902, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.8698344230651855, "learning_rate": 2.729970326409495e-07, "loss": 0.572, "mean_token_accuracy": 0.8225823640823364, "num_tokens": 24487293.0, "step": 645 }, { "epoch": 0.08217783996946953, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 5.019801139831543, "learning_rate": 2.734209410767274e-07, "loss": 0.5339, "mean_token_accuracy": 0.8304235935211182, "num_tokens": 24521403.0, "step": 646 }, { "epoch": 0.08230505024806005, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 4.91538143157959, "learning_rate": 2.738448495125053e-07, "loss": 0.5429, "mean_token_accuracy": 0.8311044573783875, "num_tokens": 24562740.0, "step": 647 }, { "epoch": 0.08243226052665055, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.975749969482422, "learning_rate": 2.7426875794828316e-07, "loss": 0.5503, "mean_token_accuracy": 0.8221328258514404, "num_tokens": 24597507.0, "step": 648 }, { "epoch": 0.08255947080524106, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.21540641784668e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.945680618286133, "learning_rate": 2.74692666384061e-07, "loss": 0.5119, "mean_token_accuracy": 0.8356629014015198, "num_tokens": 24635629.0, "step": 649 }, { "epoch": 0.08268668108383158, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.9421892166137695, "learning_rate": 2.751165748198389e-07, "loss": 0.538, "mean_token_accuracy": 0.8281950950622559, "num_tokens": 24676087.0, "step": 650 }, { "epoch": 0.08281389136242208, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.250075340270996e-06, "grad_norm": 5.076318740844727, "learning_rate": 2.755404832556168e-07, "loss": 0.556, "mean_token_accuracy": 0.8249017596244812, "num_tokens": 24713652.0, "step": 651 }, { "epoch": 0.0829411016410126, "ewc_loss": 0.0028076171875, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 4.885258197784424, "learning_rate": 2.7596439169139465e-07, "loss": 0.5382, "mean_token_accuracy": 0.830358624458313, "num_tokens": 24750130.0, "step": 652 }, { "epoch": 0.08306831191960311, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 5.001903533935547, "learning_rate": 2.763883001271725e-07, "loss": 0.5879, "mean_token_accuracy": 0.8146278858184814, "num_tokens": 24787337.0, "step": 653 }, { "epoch": 0.08319552219819361, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.304813385009766e-06, "ewc_loss_parallel": 2.250075340270996e-06, "grad_norm": 5.114404201507568, "learning_rate": 2.768122085629504e-07, "loss": 0.5798, "mean_token_accuracy": 0.8108960390090942, "num_tokens": 24822736.0, "step": 654 }, { "epoch": 0.08332273247678412, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 5.0763258934021, "learning_rate": 2.772361169987283e-07, "loss": 0.5212, "mean_token_accuracy": 0.8375897407531738, "num_tokens": 24858743.0, "step": 655 }, { "epoch": 0.08344994275537464, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.9284515380859375, "learning_rate": 2.7766002543450614e-07, "loss": 0.5297, "mean_token_accuracy": 0.8292631506919861, "num_tokens": 24901286.0, "step": 656 }, { "epoch": 0.08357715303396514, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.901594638824463, "learning_rate": 2.78083933870284e-07, "loss": 0.5991, "mean_token_accuracy": 0.8136537671089172, "num_tokens": 24943457.0, "step": 657 }, { "epoch": 0.08370436331255565, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.9726738929748535, "learning_rate": 2.785078423060619e-07, "loss": 0.5385, "mean_token_accuracy": 0.8266471028327942, "num_tokens": 24979247.0, "step": 658 }, { "epoch": 0.08383157359114617, "ewc_loss": 0.0028076171875, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 4.991944789886475, "learning_rate": 2.789317507418398e-07, "loss": 0.5512, "mean_token_accuracy": 0.8244845867156982, "num_tokens": 25017456.0, "step": 659 }, { "epoch": 0.08395878386973668, "ewc_loss": 0.0028533935546875, "ewc_loss_diag": 5.304813385009766e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 5.0381693840026855, "learning_rate": 2.7935565917761763e-07, "loss": 0.5075, "mean_token_accuracy": 0.8361110091209412, "num_tokens": 25054682.0, "step": 660 }, { "epoch": 0.08408599414832718, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.334615707397461e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 5.039157867431641, "learning_rate": 2.797795676133955e-07, "loss": 0.5807, "mean_token_accuracy": 0.8150728940963745, "num_tokens": 25092501.0, "step": 661 }, { "epoch": 0.0842132044269177, "ewc_loss": 0.0028533935546875, "ewc_loss_diag": 5.334615707397461e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 4.9488139152526855, "learning_rate": 2.802034760491734e-07, "loss": 0.4493, "mean_token_accuracy": 0.8560709953308105, "num_tokens": 25129253.0, "step": 662 }, { "epoch": 0.08434041470550821, "ewc_loss": 0.0028228759765625, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.294778823852539e-06, "grad_norm": 4.820278167724609, "learning_rate": 2.806273844849512e-07, "loss": 0.4883, "mean_token_accuracy": 0.8445833921432495, "num_tokens": 25169119.0, "step": 663 }, { "epoch": 0.08446762498409871, "ewc_loss": 0.0028076171875, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 5.024878978729248, "learning_rate": 2.810512929207291e-07, "loss": 0.559, "mean_token_accuracy": 0.8181340098381042, "num_tokens": 25207654.0, "step": 664 }, { "epoch": 0.08459483526268922, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.934897422790527, "learning_rate": 2.8147520135650697e-07, "loss": 0.504, "mean_token_accuracy": 0.8360942602157593, "num_tokens": 25243258.0, "step": 665 }, { "epoch": 0.08472204554127974, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 4.954641819000244, "learning_rate": 2.8189910979228487e-07, "loss": 0.5439, "mean_token_accuracy": 0.8292683362960815, "num_tokens": 25281301.0, "step": 666 }, { "epoch": 0.08484925581987024, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.339482307434082e-06, "grad_norm": 5.038397789001465, "learning_rate": 2.823230182280627e-07, "loss": 0.4955, "mean_token_accuracy": 0.8424244523048401, "num_tokens": 25313500.0, "step": 667 }, { "epoch": 0.08497646609846075, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3543834686279297e-06, "grad_norm": 4.909055709838867, "learning_rate": 2.827469266638406e-07, "loss": 0.5321, "mean_token_accuracy": 0.8301676511764526, "num_tokens": 25355121.0, "step": 668 }, { "epoch": 0.08510367637705127, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 4.936047077178955, "learning_rate": 2.8317083509961846e-07, "loss": 0.548, "mean_token_accuracy": 0.8280144929885864, "num_tokens": 25398868.0, "step": 669 }, { "epoch": 0.08523088665564178, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.916398525238037, "learning_rate": 2.8359474353539636e-07, "loss": 0.5271, "mean_token_accuracy": 0.831477165222168, "num_tokens": 25438670.0, "step": 670 }, { "epoch": 0.08535809693423228, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3543834686279297e-06, "grad_norm": 5.053976535797119, "learning_rate": 2.840186519711742e-07, "loss": 0.5735, "mean_token_accuracy": 0.8229905366897583, "num_tokens": 25479243.0, "step": 671 }, { "epoch": 0.0854853072128228, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 5.066756725311279, "learning_rate": 2.844425604069521e-07, "loss": 0.5579, "mean_token_accuracy": 0.8219413161277771, "num_tokens": 25514615.0, "step": 672 }, { "epoch": 0.08561251749141331, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 5.047133445739746, "learning_rate": 2.8486646884272995e-07, "loss": 0.5344, "mean_token_accuracy": 0.829345703125, "num_tokens": 25552753.0, "step": 673 }, { "epoch": 0.08573972777000381, "ewc_loss": 0.0029144287109375, "ewc_loss_diag": 5.453824996948242e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.938565254211426, "learning_rate": 2.8529037727850785e-07, "loss": 0.4718, "mean_token_accuracy": 0.8465327620506287, "num_tokens": 25587415.0, "step": 674 }, { "epoch": 0.08586693804859433, "ewc_loss": 0.0029144287109375, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.89613151550293, "learning_rate": 2.857142857142857e-07, "loss": 0.5016, "mean_token_accuracy": 0.8418324589729309, "num_tokens": 25625962.0, "step": 675 }, { "epoch": 0.08599414832718484, "ewc_loss": 0.0029144287109375, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 5.060398101806641, "learning_rate": 2.861381941500636e-07, "loss": 0.55, "mean_token_accuracy": 0.825353741645813, "num_tokens": 25669680.0, "step": 676 }, { "epoch": 0.08612135860577534, "ewc_loss": 0.0029449462890625, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 5.108506202697754, "learning_rate": 2.8656210258584144e-07, "loss": 0.5845, "mean_token_accuracy": 0.8119930624961853, "num_tokens": 25709221.0, "step": 677 }, { "epoch": 0.08624856888436586, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 5.2540740966796875, "learning_rate": 2.869860110216193e-07, "loss": 0.5855, "mean_token_accuracy": 0.8113229274749756, "num_tokens": 25741880.0, "step": 678 }, { "epoch": 0.08637577916295637, "ewc_loss": 0.0029449462890625, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 4.948330402374268, "learning_rate": 2.874099194573972e-07, "loss": 0.5186, "mean_token_accuracy": 0.8345474004745483, "num_tokens": 25786282.0, "step": 679 }, { "epoch": 0.08650298944154687, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 5.044857978820801, "learning_rate": 2.878338278931751e-07, "loss": 0.5648, "mean_token_accuracy": 0.822564959526062, "num_tokens": 25820229.0, "step": 680 }, { "epoch": 0.08663019972013739, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 5.001025199890137, "learning_rate": 2.8825773632895293e-07, "loss": 0.5, "mean_token_accuracy": 0.8439669013023376, "num_tokens": 25856465.0, "step": 681 }, { "epoch": 0.0867574099987279, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.602836608886719e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 4.963449954986572, "learning_rate": 2.886816447647308e-07, "loss": 0.4983, "mean_token_accuracy": 0.8403838276863098, "num_tokens": 25897407.0, "step": 682 }, { "epoch": 0.08688462027731841, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.602836608886719e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 4.986810684204102, "learning_rate": 2.891055532005087e-07, "loss": 0.5218, "mean_token_accuracy": 0.8338906764984131, "num_tokens": 25938825.0, "step": 683 }, { "epoch": 0.08701183055590891, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 5.116545677185059, "learning_rate": 2.8952946163628657e-07, "loss": 0.5524, "mean_token_accuracy": 0.8222293257713318, "num_tokens": 25974944.0, "step": 684 }, { "epoch": 0.08713904083449943, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.602836608886719e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 5.0231404304504395, "learning_rate": 2.899533700720644e-07, "loss": 0.5105, "mean_token_accuracy": 0.8350905179977417, "num_tokens": 26012864.0, "step": 685 }, { "epoch": 0.08726625111308994, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 4.96308708190918, "learning_rate": 2.9037727850784227e-07, "loss": 0.5565, "mean_token_accuracy": 0.8230553269386292, "num_tokens": 26049779.0, "step": 686 }, { "epoch": 0.08739346139168044, "ewc_loss": 0.0029754638671875, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 4.929200649261475, "learning_rate": 2.9080118694362016e-07, "loss": 0.5273, "mean_token_accuracy": 0.8329052329063416, "num_tokens": 26089095.0, "step": 687 }, { "epoch": 0.08752067167027096, "ewc_loss": 0.0030059814453125, "ewc_loss_diag": 5.662441253662109e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.909987926483154, "learning_rate": 2.9122509537939806e-07, "loss": 0.5267, "mean_token_accuracy": 0.8334101438522339, "num_tokens": 26132553.0, "step": 688 }, { "epoch": 0.08764788194886147, "ewc_loss": 0.0029754638671875, "ewc_loss_diag": 5.662441253662109e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 4.974364757537842, "learning_rate": 2.916490038151759e-07, "loss": 0.5775, "mean_token_accuracy": 0.8139150142669678, "num_tokens": 26171301.0, "step": 689 }, { "epoch": 0.08777509222745197, "ewc_loss": 0.0030059814453125, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 5.0999836921691895, "learning_rate": 2.9207291225095376e-07, "loss": 0.5647, "mean_token_accuracy": 0.8196671009063721, "num_tokens": 26217871.0, "step": 690 }, { "epoch": 0.08790230250604249, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.7220458984375e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.939024448394775, "learning_rate": 2.9249682068673166e-07, "loss": 0.5285, "mean_token_accuracy": 0.8304450511932373, "num_tokens": 26255927.0, "step": 691 }, { "epoch": 0.088029512784633, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.7220458984375e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 4.938717365264893, "learning_rate": 2.9292072912250955e-07, "loss": 0.5113, "mean_token_accuracy": 0.8388336896896362, "num_tokens": 26296041.0, "step": 692 }, { "epoch": 0.0881567230632235, "ewc_loss": 0.0030059814453125, "ewc_loss_diag": 5.751848220825195e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 5.115733623504639, "learning_rate": 2.933446375582874e-07, "loss": 0.5937, "mean_token_accuracy": 0.8130251169204712, "num_tokens": 26333785.0, "step": 693 }, { "epoch": 0.08828393334181402, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.036827564239502, "learning_rate": 2.9376854599406525e-07, "loss": 0.5003, "mean_token_accuracy": 0.8373428583145142, "num_tokens": 26368837.0, "step": 694 }, { "epoch": 0.08841114362040453, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 5.0991082191467285, "learning_rate": 2.9419245442984315e-07, "loss": 0.5747, "mean_token_accuracy": 0.8179076910018921, "num_tokens": 26402151.0, "step": 695 }, { "epoch": 0.08853835389899505, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 5.010698318481445, "learning_rate": 2.9461636286562104e-07, "loss": 0.5278, "mean_token_accuracy": 0.8305874466896057, "num_tokens": 26435550.0, "step": 696 }, { "epoch": 0.08866556417758555, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.944799423217773, "learning_rate": 2.9504027130139884e-07, "loss": 0.5178, "mean_token_accuracy": 0.8356771469116211, "num_tokens": 26479299.0, "step": 697 }, { "epoch": 0.08879277445617606, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.988734722137451, "learning_rate": 2.9546417973717674e-07, "loss": 0.554, "mean_token_accuracy": 0.8239974975585938, "num_tokens": 26519380.0, "step": 698 }, { "epoch": 0.08891998473476657, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.998802661895752, "learning_rate": 2.9588808817295464e-07, "loss": 0.4801, "mean_token_accuracy": 0.8440855145454407, "num_tokens": 26557306.0, "step": 699 }, { "epoch": 0.08904719501335707, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 5.085526466369629, "learning_rate": 2.9631199660873253e-07, "loss": 0.5502, "mean_token_accuracy": 0.8261067867279053, "num_tokens": 26593549.0, "step": 700 }, { "epoch": 0.08917440529194759, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.982293605804443, "learning_rate": 2.9673590504451033e-07, "loss": 0.475, "mean_token_accuracy": 0.8485350608825684, "num_tokens": 26629926.0, "step": 701 }, { "epoch": 0.0893016155705381, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.990947723388672, "learning_rate": 2.9715981348028823e-07, "loss": 0.5069, "mean_token_accuracy": 0.8394337892532349, "num_tokens": 26668652.0, "step": 702 }, { "epoch": 0.0894288258491286, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.199102401733398, "learning_rate": 2.975837219160661e-07, "loss": 0.5691, "mean_token_accuracy": 0.819014310836792, "num_tokens": 26708084.0, "step": 703 }, { "epoch": 0.08955603612771912, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.014420986175537, "learning_rate": 2.98007630351844e-07, "loss": 0.5351, "mean_token_accuracy": 0.830464780330658, "num_tokens": 26750134.0, "step": 704 }, { "epoch": 0.08968324640630963, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 5.192662239074707, "learning_rate": 2.984315387876218e-07, "loss": 0.5435, "mean_token_accuracy": 0.82627272605896, "num_tokens": 26785410.0, "step": 705 }, { "epoch": 0.08981045668490013, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.136507034301758, "learning_rate": 2.988554472233997e-07, "loss": 0.548, "mean_token_accuracy": 0.8286446332931519, "num_tokens": 26825226.0, "step": 706 }, { "epoch": 0.08993766696349065, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 4.982025623321533, "learning_rate": 2.992793556591776e-07, "loss": 0.5409, "mean_token_accuracy": 0.8273565769195557, "num_tokens": 26865401.0, "step": 707 }, { "epoch": 0.09006487724208116, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 5.009066104888916, "learning_rate": 2.997032640949555e-07, "loss": 0.5512, "mean_token_accuracy": 0.8258838653564453, "num_tokens": 26905365.0, "step": 708 }, { "epoch": 0.09019208752067168, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.02521276473999, "learning_rate": 3.001271725307333e-07, "loss": 0.4917, "mean_token_accuracy": 0.8418100476264954, "num_tokens": 26941348.0, "step": 709 }, { "epoch": 0.09031929779926218, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.9604644775390625e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.189945220947266, "learning_rate": 3.005510809665112e-07, "loss": 0.5278, "mean_token_accuracy": 0.8321752548217773, "num_tokens": 26975803.0, "step": 710 }, { "epoch": 0.09044650807785269, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.9604644775390625e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.076581954956055, "learning_rate": 3.009749894022891e-07, "loss": 0.4733, "mean_token_accuracy": 0.8486409187316895, "num_tokens": 27014931.0, "step": 711 }, { "epoch": 0.0905737183564432, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.004902362823486, "learning_rate": 3.01398897838067e-07, "loss": 0.5294, "mean_token_accuracy": 0.8336699604988098, "num_tokens": 27055505.0, "step": 712 }, { "epoch": 0.0907009286350337, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.069647312164307, "learning_rate": 3.018228062738448e-07, "loss": 0.5573, "mean_token_accuracy": 0.8261622190475464, "num_tokens": 27092121.0, "step": 713 }, { "epoch": 0.09082813891362422, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.061403274536133, "learning_rate": 3.022467147096227e-07, "loss": 0.4945, "mean_token_accuracy": 0.8422199487686157, "num_tokens": 27129827.0, "step": 714 }, { "epoch": 0.09095534919221474, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 4.991945743560791, "learning_rate": 3.026706231454006e-07, "loss": 0.5111, "mean_token_accuracy": 0.8356810808181763, "num_tokens": 27174977.0, "step": 715 }, { "epoch": 0.09108255947080524, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.154153823852539, "learning_rate": 3.0309453158117844e-07, "loss": 0.5171, "mean_token_accuracy": 0.8345866203308105, "num_tokens": 27211472.0, "step": 716 }, { "epoch": 0.09120976974939575, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.1176838874816895, "learning_rate": 3.035184400169563e-07, "loss": 0.5983, "mean_token_accuracy": 0.8139205574989319, "num_tokens": 27249160.0, "step": 717 }, { "epoch": 0.09133698002798626, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.1570329666137695, "learning_rate": 3.039423484527342e-07, "loss": 0.5085, "mean_token_accuracy": 0.8360345959663391, "num_tokens": 27287946.0, "step": 718 }, { "epoch": 0.09146419030657676, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.082759857177734, "learning_rate": 3.043662568885121e-07, "loss": 0.5413, "mean_token_accuracy": 0.8229634761810303, "num_tokens": 27323388.0, "step": 719 }, { "epoch": 0.09159140058516728, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.234404563903809, "learning_rate": 3.0479016532428993e-07, "loss": 0.5901, "mean_token_accuracy": 0.8123751878738403, "num_tokens": 27357524.0, "step": 720 }, { "epoch": 0.0917186108637578, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.08915376663208, "learning_rate": 3.052140737600678e-07, "loss": 0.4967, "mean_token_accuracy": 0.840083122253418, "num_tokens": 27392037.0, "step": 721 }, { "epoch": 0.09184582114234831, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.0411882400512695, "learning_rate": 3.056379821958457e-07, "loss": 0.5098, "mean_token_accuracy": 0.8380917310714722, "num_tokens": 27429149.0, "step": 722 }, { "epoch": 0.09197303142093881, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.112454414367676, "learning_rate": 3.060618906316236e-07, "loss": 0.513, "mean_token_accuracy": 0.8386833071708679, "num_tokens": 27465023.0, "step": 723 }, { "epoch": 0.09210024169952932, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.101386547088623, "learning_rate": 3.064857990674014e-07, "loss": 0.505, "mean_token_accuracy": 0.8388255834579468, "num_tokens": 27501268.0, "step": 724 }, { "epoch": 0.09222745197811984, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.135283946990967, "learning_rate": 3.0690970750317927e-07, "loss": 0.5569, "mean_token_accuracy": 0.8241620063781738, "num_tokens": 27539937.0, "step": 725 }, { "epoch": 0.09235466225671034, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.175310134887695, "learning_rate": 3.0733361593895717e-07, "loss": 0.5191, "mean_token_accuracy": 0.8339117765426636, "num_tokens": 27574576.0, "step": 726 }, { "epoch": 0.09248187253530085, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.147214412689209, "learning_rate": 3.0775752437473507e-07, "loss": 0.5276, "mean_token_accuracy": 0.8279567956924438, "num_tokens": 27616036.0, "step": 727 }, { "epoch": 0.09260908281389137, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.043818950653076, "learning_rate": 3.081814328105129e-07, "loss": 0.5507, "mean_token_accuracy": 0.8248624801635742, "num_tokens": 27662040.0, "step": 728 }, { "epoch": 0.09273629309248187, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.115675449371338, "learning_rate": 3.0860534124629076e-07, "loss": 0.5798, "mean_token_accuracy": 0.8159024119377136, "num_tokens": 27705136.0, "step": 729 }, { "epoch": 0.09286350337107238, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.219296932220459, "learning_rate": 3.0902924968206866e-07, "loss": 0.5296, "mean_token_accuracy": 0.8319469690322876, "num_tokens": 27740564.0, "step": 730 }, { "epoch": 0.0929907136496629, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.240065097808838, "learning_rate": 3.0945315811784656e-07, "loss": 0.5027, "mean_token_accuracy": 0.8368362188339233, "num_tokens": 27774788.0, "step": 731 }, { "epoch": 0.0931179239282534, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.043883800506592, "learning_rate": 3.098770665536244e-07, "loss": 0.6473, "mean_token_accuracy": 0.7932852506637573, "num_tokens": 27818300.0, "step": 732 }, { "epoch": 0.09324513420684391, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.128617763519287, "learning_rate": 3.1030097498940225e-07, "loss": 0.4907, "mean_token_accuracy": 0.8420649766921997, "num_tokens": 27857739.0, "step": 733 }, { "epoch": 0.09337234448543442, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.184666156768799, "learning_rate": 3.1072488342518015e-07, "loss": 0.5292, "mean_token_accuracy": 0.825745701789856, "num_tokens": 27893897.0, "step": 734 }, { "epoch": 0.09349955476402494, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 2.5331974029541016e-06, "grad_norm": 5.065469741821289, "learning_rate": 3.11148791860958e-07, "loss": 0.5161, "mean_token_accuracy": 0.8329504728317261, "num_tokens": 27930511.0, "step": 735 }, { "epoch": 0.09362676504261544, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.251402854919434, "learning_rate": 3.115727002967359e-07, "loss": 0.5339, "mean_token_accuracy": 0.8293484449386597, "num_tokens": 27961128.0, "step": 736 }, { "epoch": 0.09375397532120595, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 2.5331974029541016e-06, "grad_norm": 5.123168468475342, "learning_rate": 3.1199660873251374e-07, "loss": 0.5342, "mean_token_accuracy": 0.8306814432144165, "num_tokens": 28000748.0, "step": 737 }, { "epoch": 0.09388118559979647, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.043169021606445, "learning_rate": 3.1242051716829164e-07, "loss": 0.5678, "mean_token_accuracy": 0.8180937170982361, "num_tokens": 28046455.0, "step": 738 }, { "epoch": 0.09400839587838697, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 2.5331974029541016e-06, "grad_norm": 5.163820743560791, "learning_rate": 3.128444256040695e-07, "loss": 0.5438, "mean_token_accuracy": 0.828211784362793, "num_tokens": 28087026.0, "step": 739 }, { "epoch": 0.09413560615697748, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.073704719543457, "learning_rate": 3.132683340398474e-07, "loss": 0.4672, "mean_token_accuracy": 0.8481528759002686, "num_tokens": 28126382.0, "step": 740 }, { "epoch": 0.094262816435568, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.1119279861450195, "learning_rate": 3.1369224247562523e-07, "loss": 0.4935, "mean_token_accuracy": 0.8417817950248718, "num_tokens": 28161198.0, "step": 741 }, { "epoch": 0.0943900267141585, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.22868537902832e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.13208532333374, "learning_rate": 3.1411615091140313e-07, "loss": 0.4627, "mean_token_accuracy": 0.8512048721313477, "num_tokens": 28204612.0, "step": 742 }, { "epoch": 0.09451723699274901, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.22868537902832e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.093149185180664, "learning_rate": 3.14540059347181e-07, "loss": 0.5168, "mean_token_accuracy": 0.8348870873451233, "num_tokens": 28241050.0, "step": 743 }, { "epoch": 0.09464444727133953, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.144480228424072, "learning_rate": 3.149639677829589e-07, "loss": 0.4773, "mean_token_accuracy": 0.8447117805480957, "num_tokens": 28278032.0, "step": 744 }, { "epoch": 0.09477165754993004, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.318092346191406e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.199569225311279, "learning_rate": 3.153878762187368e-07, "loss": 0.5642, "mean_token_accuracy": 0.82524573802948, "num_tokens": 28315599.0, "step": 745 }, { "epoch": 0.09489886782852054, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.096126556396484, "learning_rate": 3.158117846545146e-07, "loss": 0.5882, "mean_token_accuracy": 0.8157957792282104, "num_tokens": 28358504.0, "step": 746 }, { "epoch": 0.09502607810711106, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.245214939117432, "learning_rate": 3.1623569309029247e-07, "loss": 0.5046, "mean_token_accuracy": 0.8384485244750977, "num_tokens": 28389626.0, "step": 747 }, { "epoch": 0.09515328838570157, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.087305545806885, "learning_rate": 3.1665960152607037e-07, "loss": 0.4867, "mean_token_accuracy": 0.8421218395233154, "num_tokens": 28430361.0, "step": 748 }, { "epoch": 0.09528049866429207, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.5331974029541016e-06, "grad_norm": 5.061094760894775, "learning_rate": 3.1708350996184826e-07, "loss": 0.5238, "mean_token_accuracy": 0.8317068219184875, "num_tokens": 28469792.0, "step": 749 }, { "epoch": 0.09540770894288259, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.155885219573975, "learning_rate": 3.175074183976261e-07, "loss": 0.5026, "mean_token_accuracy": 0.841444730758667, "num_tokens": 28507569.0, "step": 750 }, { "epoch": 0.0955349192214731, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.209061622619629, "learning_rate": 3.1793132683340396e-07, "loss": 0.5285, "mean_token_accuracy": 0.8321065902709961, "num_tokens": 28542336.0, "step": 751 }, { "epoch": 0.0956621295000636, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.174620628356934, "learning_rate": 3.1835523526918186e-07, "loss": 0.4786, "mean_token_accuracy": 0.8448936343193054, "num_tokens": 28578587.0, "step": 752 }, { "epoch": 0.09578933977865411, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.147195339202881, "learning_rate": 3.1877914370495975e-07, "loss": 0.5584, "mean_token_accuracy": 0.8227939605712891, "num_tokens": 28617505.0, "step": 753 }, { "epoch": 0.09591655005724463, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 2.5331974029541016e-06, "grad_norm": 5.191051006317139, "learning_rate": 3.1920305214073755e-07, "loss": 0.5638, "mean_token_accuracy": 0.8225467205047607, "num_tokens": 28658526.0, "step": 754 }, { "epoch": 0.09604376033583513, "ewc_loss": 0.0032196044921875, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.137179374694824, "learning_rate": 3.1962696057651545e-07, "loss": 0.4997, "mean_token_accuracy": 0.83969646692276, "num_tokens": 28698825.0, "step": 755 }, { "epoch": 0.09617097061442564, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.2101359367370605, "learning_rate": 3.2005086901229335e-07, "loss": 0.5425, "mean_token_accuracy": 0.824319064617157, "num_tokens": 28733238.0, "step": 756 }, { "epoch": 0.09629818089301616, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.186224937438965, "learning_rate": 3.2047477744807125e-07, "loss": 0.5436, "mean_token_accuracy": 0.8297428488731384, "num_tokens": 28771552.0, "step": 757 }, { "epoch": 0.09642539117160667, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.592802047729492e-06, "grad_norm": 5.11186408996582, "learning_rate": 3.2089868588384904e-07, "loss": 0.5742, "mean_token_accuracy": 0.8154717087745667, "num_tokens": 28810827.0, "step": 758 }, { "epoch": 0.09655260145019717, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.121541500091553, "learning_rate": 3.2132259431962694e-07, "loss": 0.4877, "mean_token_accuracy": 0.8442659974098206, "num_tokens": 28849017.0, "step": 759 }, { "epoch": 0.09667981172878769, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.220428466796875, "learning_rate": 3.2174650275540484e-07, "loss": 0.5378, "mean_token_accuracy": 0.8340966701507568, "num_tokens": 28887659.0, "step": 760 }, { "epoch": 0.0968070220073782, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.11582088470459, "learning_rate": 3.2217041119118274e-07, "loss": 0.5038, "mean_token_accuracy": 0.8382845520973206, "num_tokens": 28928983.0, "step": 761 }, { "epoch": 0.0969342322859687, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.592802047729492e-06, "grad_norm": 5.205071449279785, "learning_rate": 3.2259431962696053e-07, "loss": 0.5367, "mean_token_accuracy": 0.8299921751022339, "num_tokens": 28962195.0, "step": 762 }, { "epoch": 0.09706144256455922, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.208623886108398, "learning_rate": 3.2301822806273843e-07, "loss": 0.5194, "mean_token_accuracy": 0.8317987322807312, "num_tokens": 28997351.0, "step": 763 }, { "epoch": 0.09718865284314973, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.169093608856201, "learning_rate": 3.2344213649851633e-07, "loss": 0.5456, "mean_token_accuracy": 0.8263561129570007, "num_tokens": 29040988.0, "step": 764 }, { "epoch": 0.09731586312174023, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.1523847579956055, "learning_rate": 3.238660449342942e-07, "loss": 0.5081, "mean_token_accuracy": 0.8350226879119873, "num_tokens": 29081873.0, "step": 765 }, { "epoch": 0.09744307340033075, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.22387170791626, "learning_rate": 3.24289953370072e-07, "loss": 0.4799, "mean_token_accuracy": 0.8425459265708923, "num_tokens": 29115360.0, "step": 766 }, { "epoch": 0.09757028367892126, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.1090617179870605, "learning_rate": 3.247138618058499e-07, "loss": 0.5248, "mean_token_accuracy": 0.8273491263389587, "num_tokens": 29160139.0, "step": 767 }, { "epoch": 0.09769749395751176, "ewc_loss": 0.0032806396484375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.180042266845703, "learning_rate": 3.251377702416278e-07, "loss": 0.5542, "mean_token_accuracy": 0.8246800899505615, "num_tokens": 29202872.0, "step": 768 }, { "epoch": 0.09782470423610228, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.164124965667725, "learning_rate": 3.255616786774057e-07, "loss": 0.5016, "mean_token_accuracy": 0.8384679555892944, "num_tokens": 29246965.0, "step": 769 }, { "epoch": 0.09795191451469279, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.106716156005859, "learning_rate": 3.259855871131835e-07, "loss": 0.4413, "mean_token_accuracy": 0.8556501865386963, "num_tokens": 29289531.0, "step": 770 }, { "epoch": 0.0980791247932833, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.137127876281738, "learning_rate": 3.264094955489614e-07, "loss": 0.498, "mean_token_accuracy": 0.8398380875587463, "num_tokens": 29330484.0, "step": 771 }, { "epoch": 0.0982063350718738, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.173183917999268, "learning_rate": 3.268334039847393e-07, "loss": 0.5245, "mean_token_accuracy": 0.8306430578231812, "num_tokens": 29369595.0, "step": 772 }, { "epoch": 0.09833354535046432, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.159115791320801, "learning_rate": 3.2725731242051715e-07, "loss": 0.5131, "mean_token_accuracy": 0.838451623916626, "num_tokens": 29411856.0, "step": 773 }, { "epoch": 0.09846075562905483, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.127908229827881, "learning_rate": 3.27681220856295e-07, "loss": 0.4958, "mean_token_accuracy": 0.841957688331604, "num_tokens": 29454754.0, "step": 774 }, { "epoch": 0.09858796590764533, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.1986188888549805, "learning_rate": 3.281051292920729e-07, "loss": 0.4912, "mean_token_accuracy": 0.8396201729774475, "num_tokens": 29489131.0, "step": 775 }, { "epoch": 0.09871517618623585, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.173831462860107, "learning_rate": 3.285290377278508e-07, "loss": 0.546, "mean_token_accuracy": 0.823442816734314, "num_tokens": 29528292.0, "step": 776 }, { "epoch": 0.09884238646482636, "ewc_loss": 0.0032806396484375, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.180777072906494, "learning_rate": 3.2895294616362864e-07, "loss": 0.4791, "mean_token_accuracy": 0.8465595245361328, "num_tokens": 29564878.0, "step": 777 }, { "epoch": 0.09896959674341686, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.155076026916504, "learning_rate": 3.293768545994065e-07, "loss": 0.4985, "mean_token_accuracy": 0.8377212285995483, "num_tokens": 29602826.0, "step": 778 }, { "epoch": 0.09909680702200738, "ewc_loss": 0.0033111572265625, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.2090044021606445, "learning_rate": 3.298007630351844e-07, "loss": 0.5669, "mean_token_accuracy": 0.8216322064399719, "num_tokens": 29640323.0, "step": 779 }, { "epoch": 0.09922401730059789, "ewc_loss": 0.0033111572265625, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.125085830688477, "learning_rate": 3.302246714709623e-07, "loss": 0.4969, "mean_token_accuracy": 0.8393036127090454, "num_tokens": 29683302.0, "step": 780 }, { "epoch": 0.09935122757918839, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.233211517333984, "learning_rate": 3.3064857990674013e-07, "loss": 0.5215, "mean_token_accuracy": 0.836424708366394, "num_tokens": 29721085.0, "step": 781 }, { "epoch": 0.0994784378577789, "ewc_loss": 0.0033416748046875, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 5.261393070220947, "learning_rate": 3.31072488342518e-07, "loss": 0.4737, "mean_token_accuracy": 0.8458724021911621, "num_tokens": 29753810.0, "step": 782 }, { "epoch": 0.09960564813636942, "ewc_loss": 0.0033416748046875, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 5.1169281005859375, "learning_rate": 3.314963967782959e-07, "loss": 0.5014, "mean_token_accuracy": 0.8396807909011841, "num_tokens": 29796488.0, "step": 783 }, { "epoch": 0.09973285841495994, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.169332027435303, "learning_rate": 3.319203052140738e-07, "loss": 0.5114, "mean_token_accuracy": 0.8351014852523804, "num_tokens": 29838181.0, "step": 784 }, { "epoch": 0.09986006869355044, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.407943248748779, "learning_rate": 3.323442136498516e-07, "loss": 0.4935, "mean_token_accuracy": 0.8426196575164795, "num_tokens": 29868893.0, "step": 785 }, { "epoch": 0.09998727897214095, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.705522537231445e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.311955451965332, "learning_rate": 3.3276812208562947e-07, "loss": 0.509, "mean_token_accuracy": 0.8347259759902954, "num_tokens": 29903571.0, "step": 786 }, { "epoch": 0.10011448925073146, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.705522537231445e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.239852428436279, "learning_rate": 3.3319203052140737e-07, "loss": 0.5663, "mean_token_accuracy": 0.826316773891449, "num_tokens": 29940438.0, "step": 787 }, { "epoch": 0.10024169952932196, "ewc_loss": 0.0033111572265625, "ewc_loss_diag": 6.705522537231445e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.234591484069824, "learning_rate": 3.336159389571852e-07, "loss": 0.4885, "mean_token_accuracy": 0.8465942144393921, "num_tokens": 29976538.0, "step": 788 }, { "epoch": 0.10036890980791248, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.167147636413574, "learning_rate": 3.340398473929631e-07, "loss": 0.5589, "mean_token_accuracy": 0.8225648403167725, "num_tokens": 30019814.0, "step": 789 }, { "epoch": 0.100496120086503, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.252498149871826, "learning_rate": 3.3446375582874096e-07, "loss": 0.5085, "mean_token_accuracy": 0.8377916812896729, "num_tokens": 30057578.0, "step": 790 }, { "epoch": 0.1006233303650935, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.252221584320068, "learning_rate": 3.3488766426451886e-07, "loss": 0.5808, "mean_token_accuracy": 0.8235982060432434, "num_tokens": 30094869.0, "step": 791 }, { "epoch": 0.10075054064368401, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.200406074523926, "learning_rate": 3.353115727002967e-07, "loss": 0.5419, "mean_token_accuracy": 0.8284756541252136, "num_tokens": 30136446.0, "step": 792 }, { "epoch": 0.10087775092227452, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.183835029602051, "learning_rate": 3.357354811360746e-07, "loss": 0.5357, "mean_token_accuracy": 0.8247539401054382, "num_tokens": 30179226.0, "step": 793 }, { "epoch": 0.10100496120086502, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.291323661804199, "learning_rate": 3.3615938957185245e-07, "loss": 0.5267, "mean_token_accuracy": 0.8311412930488586, "num_tokens": 30213890.0, "step": 794 }, { "epoch": 0.10113217147945554, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.292294025421143, "learning_rate": 3.3658329800763035e-07, "loss": 0.5765, "mean_token_accuracy": 0.8136521577835083, "num_tokens": 30245691.0, "step": 795 }, { "epoch": 0.10125938175804605, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.234254837036133, "learning_rate": 3.370072064434082e-07, "loss": 0.5319, "mean_token_accuracy": 0.8223769664764404, "num_tokens": 30281738.0, "step": 796 }, { "epoch": 0.10138659203663657, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.176501274108887, "learning_rate": 3.374311148791861e-07, "loss": 0.4965, "mean_token_accuracy": 0.8386751413345337, "num_tokens": 30320707.0, "step": 797 }, { "epoch": 0.10151380231522707, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.379044532775879, "learning_rate": 3.3785502331496394e-07, "loss": 0.5046, "mean_token_accuracy": 0.8364675045013428, "num_tokens": 30351379.0, "step": 798 }, { "epoch": 0.10164101259381758, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.379429817199707, "learning_rate": 3.3827893175074184e-07, "loss": 0.5286, "mean_token_accuracy": 0.8289178609848022, "num_tokens": 30382395.0, "step": 799 }, { "epoch": 0.1017682228724081, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 5.249699592590332, "learning_rate": 3.387028401865197e-07, "loss": 0.5346, "mean_token_accuracy": 0.8330780863761902, "num_tokens": 30419758.0, "step": 800 }, { "epoch": 0.1018954331509986, "ewc_loss": 0.00341796875, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 5.254354953765869, "learning_rate": 3.391267486222976e-07, "loss": 0.5187, "mean_token_accuracy": 0.8309859037399292, "num_tokens": 30458115.0, "step": 801 }, { "epoch": 0.10202264342958911, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 5.401701927185059, "learning_rate": 3.3955065705807543e-07, "loss": 0.553, "mean_token_accuracy": 0.8211112022399902, "num_tokens": 30489679.0, "step": 802 }, { "epoch": 0.10214985370817962, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.279511451721191, "learning_rate": 3.3997456549385333e-07, "loss": 0.4897, "mean_token_accuracy": 0.8430538773536682, "num_tokens": 30524874.0, "step": 803 }, { "epoch": 0.10227706398677013, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.290587425231934, "learning_rate": 3.403984739296312e-07, "loss": 0.5631, "mean_token_accuracy": 0.821398913860321, "num_tokens": 30558354.0, "step": 804 }, { "epoch": 0.10240427426536064, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 5.233362197875977, "learning_rate": 3.408223823654091e-07, "loss": 0.494, "mean_token_accuracy": 0.840271532535553, "num_tokens": 30594827.0, "step": 805 }, { "epoch": 0.10253148454395115, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.2852783203125, "learning_rate": 3.412462908011869e-07, "loss": 0.5351, "mean_token_accuracy": 0.8349866271018982, "num_tokens": 30635259.0, "step": 806 }, { "epoch": 0.10265869482254165, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.250602722167969, "learning_rate": 3.4167019923696477e-07, "loss": 0.5279, "mean_token_accuracy": 0.8363394737243652, "num_tokens": 30681850.0, "step": 807 }, { "epoch": 0.10278590510113217, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.274986743927002, "learning_rate": 3.4209410767274267e-07, "loss": 0.4946, "mean_token_accuracy": 0.8451704382896423, "num_tokens": 30721302.0, "step": 808 }, { "epoch": 0.10291311537972268, "ewc_loss": 0.00347900390625, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 5.355229377746582, "learning_rate": 3.4251801610852057e-07, "loss": 0.5693, "mean_token_accuracy": 0.8161023855209351, "num_tokens": 30761975.0, "step": 809 }, { "epoch": 0.1030403256583132, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.344712257385254, "learning_rate": 3.429419245442984e-07, "loss": 0.5217, "mean_token_accuracy": 0.8311296105384827, "num_tokens": 30797754.0, "step": 810 }, { "epoch": 0.1031675359369037, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.231648921966553, "learning_rate": 3.4336583298007626e-07, "loss": 0.496, "mean_token_accuracy": 0.8418232202529907, "num_tokens": 30834183.0, "step": 811 }, { "epoch": 0.10329474621549421, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.211845874786377, "learning_rate": 3.4378974141585416e-07, "loss": 0.523, "mean_token_accuracy": 0.8339883089065552, "num_tokens": 30874160.0, "step": 812 }, { "epoch": 0.10342195649408473, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.251540184020996, "learning_rate": 3.4421364985163206e-07, "loss": 0.4614, "mean_token_accuracy": 0.8496939539909363, "num_tokens": 30916827.0, "step": 813 }, { "epoch": 0.10354916677267523, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.272741794586182, "learning_rate": 3.446375582874099e-07, "loss": 0.5325, "mean_token_accuracy": 0.8319613933563232, "num_tokens": 30960063.0, "step": 814 }, { "epoch": 0.10367637705126574, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.322780609130859, "learning_rate": 3.4506146672318775e-07, "loss": 0.4796, "mean_token_accuracy": 0.843002438545227, "num_tokens": 30995172.0, "step": 815 }, { "epoch": 0.10380358732985626, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.308996200561523, "learning_rate": 3.4548537515896565e-07, "loss": 0.5396, "mean_token_accuracy": 0.827613115310669, "num_tokens": 31035089.0, "step": 816 }, { "epoch": 0.10393079760844676, "ewc_loss": 0.00347900390625, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 5.248507499694824, "learning_rate": 3.4590928359474355e-07, "loss": 0.5218, "mean_token_accuracy": 0.8318371772766113, "num_tokens": 31074836.0, "step": 817 }, { "epoch": 0.10405800788703727, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.23354959487915, "learning_rate": 3.463331920305214e-07, "loss": 0.541, "mean_token_accuracy": 0.8262351751327515, "num_tokens": 31114700.0, "step": 818 }, { "epoch": 0.10418521816562779, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.3631205558776855, "learning_rate": 3.4675710046629924e-07, "loss": 0.4892, "mean_token_accuracy": 0.8418698310852051, "num_tokens": 31148258.0, "step": 819 }, { "epoch": 0.1043124284442183, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.336245059967041, "learning_rate": 3.4718100890207714e-07, "loss": 0.5287, "mean_token_accuracy": 0.8313835859298706, "num_tokens": 31185720.0, "step": 820 }, { "epoch": 0.1044396387228088, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.3288164138793945, "learning_rate": 3.4760491733785504e-07, "loss": 0.5391, "mean_token_accuracy": 0.8244733810424805, "num_tokens": 31220353.0, "step": 821 }, { "epoch": 0.10456684900139931, "ewc_loss": 0.0035247802734375, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.326324939727783, "learning_rate": 3.480288257736329e-07, "loss": 0.5163, "mean_token_accuracy": 0.831524133682251, "num_tokens": 31256477.0, "step": 822 }, { "epoch": 0.10469405927998983, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.289747714996338, "learning_rate": 3.4845273420941073e-07, "loss": 0.5152, "mean_token_accuracy": 0.8364397287368774, "num_tokens": 31294365.0, "step": 823 }, { "epoch": 0.10482126955858033, "ewc_loss": 0.0035247802734375, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.449944496154785, "learning_rate": 3.4887664264518863e-07, "loss": 0.5105, "mean_token_accuracy": 0.8329225778579712, "num_tokens": 31329357.0, "step": 824 }, { "epoch": 0.10494847983717084, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 5.254401206970215, "learning_rate": 3.4930055108096653e-07, "loss": 0.5157, "mean_token_accuracy": 0.8345515727996826, "num_tokens": 31372180.0, "step": 825 }, { "epoch": 0.10507569011576136, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 7.033348083496094e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 5.238521575927734, "learning_rate": 3.497244595167443e-07, "loss": 0.4989, "mean_token_accuracy": 0.8390056490898132, "num_tokens": 31414866.0, "step": 826 }, { "epoch": 0.10520290039435186, "ewc_loss": 0.0035247802734375, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.416589260101318, "learning_rate": 3.501483679525222e-07, "loss": 0.5674, "mean_token_accuracy": 0.8189482092857361, "num_tokens": 31450583.0, "step": 827 }, { "epoch": 0.10533011067294237, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.301612377166748, "learning_rate": 3.505722763883001e-07, "loss": 0.4989, "mean_token_accuracy": 0.8375589847564697, "num_tokens": 31490371.0, "step": 828 }, { "epoch": 0.10545732095153289, "ewc_loss": 0.0035247802734375, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.3127970695495605, "learning_rate": 3.50996184824078e-07, "loss": 0.5408, "mean_token_accuracy": 0.8262214064598083, "num_tokens": 31525951.0, "step": 829 }, { "epoch": 0.10558453123012339, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 5.260552883148193, "learning_rate": 3.514200932598558e-07, "loss": 0.5802, "mean_token_accuracy": 0.8165103197097778, "num_tokens": 31570109.0, "step": 830 }, { "epoch": 0.1057117415087139, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.354426860809326, "learning_rate": 3.518440016956337e-07, "loss": 0.5227, "mean_token_accuracy": 0.8324301242828369, "num_tokens": 31610569.0, "step": 831 }, { "epoch": 0.10583895178730442, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.299647331237793, "learning_rate": 3.522679101314116e-07, "loss": 0.5348, "mean_token_accuracy": 0.8320541381835938, "num_tokens": 31653585.0, "step": 832 }, { "epoch": 0.10596616206589493, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.277735710144043, "learning_rate": 3.526918185671895e-07, "loss": 0.5444, "mean_token_accuracy": 0.8255562782287598, "num_tokens": 31696353.0, "step": 833 }, { "epoch": 0.10609337234448543, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.357211589813232, "learning_rate": 3.531157270029673e-07, "loss": 0.4755, "mean_token_accuracy": 0.8475085496902466, "num_tokens": 31730620.0, "step": 834 }, { "epoch": 0.10622058262307595, "ewc_loss": 0.003631591796875, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.295902729034424, "learning_rate": 3.535396354387452e-07, "loss": 0.4914, "mean_token_accuracy": 0.8402745723724365, "num_tokens": 31770538.0, "step": 835 }, { "epoch": 0.10634779290166646, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 2.8461217880249023e-06, "grad_norm": 5.335759162902832, "learning_rate": 3.539635438745231e-07, "loss": 0.5336, "mean_token_accuracy": 0.8289400339126587, "num_tokens": 31808687.0, "step": 836 }, { "epoch": 0.10647500318025696, "ewc_loss": 0.003631591796875, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.283992290496826, "learning_rate": 3.54387452310301e-07, "loss": 0.5128, "mean_token_accuracy": 0.8350387811660767, "num_tokens": 31852310.0, "step": 837 }, { "epoch": 0.10660221345884748, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.313827037811279, "learning_rate": 3.548113607460788e-07, "loss": 0.4814, "mean_token_accuracy": 0.8479255437850952, "num_tokens": 31887897.0, "step": 838 }, { "epoch": 0.10672942373743799, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.324522495269775, "learning_rate": 3.552352691818567e-07, "loss": 0.5068, "mean_token_accuracy": 0.8387909531593323, "num_tokens": 31926159.0, "step": 839 }, { "epoch": 0.10685663401602849, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 2.9355287551879883e-06, "grad_norm": 5.310186862945557, "learning_rate": 3.556591776176346e-07, "loss": 0.481, "mean_token_accuracy": 0.8457925319671631, "num_tokens": 31968432.0, "step": 840 }, { "epoch": 0.106983844294619, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 2.9355287551879883e-06, "grad_norm": 5.444575786590576, "learning_rate": 3.560830860534125e-07, "loss": 0.5145, "mean_token_accuracy": 0.8374053239822388, "num_tokens": 32000949.0, "step": 841 }, { "epoch": 0.10711105457320952, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.435650825500488, "learning_rate": 3.565069944891903e-07, "loss": 0.557, "mean_token_accuracy": 0.8216400146484375, "num_tokens": 32033862.0, "step": 842 }, { "epoch": 0.10723826485180002, "ewc_loss": 0.003662109375, "ewc_loss_diag": 7.241964340209961e-06, "ewc_loss_parallel": 2.9355287551879883e-06, "grad_norm": 5.383221626281738, "learning_rate": 3.569309029249682e-07, "loss": 0.5192, "mean_token_accuracy": 0.8351308703422546, "num_tokens": 32068623.0, "step": 843 }, { "epoch": 0.10736547513039053, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.241964340209961e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.386567115783691, "learning_rate": 3.573548113607461e-07, "loss": 0.5059, "mean_token_accuracy": 0.8393406867980957, "num_tokens": 32106658.0, "step": 844 }, { "epoch": 0.10749268540898105, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.361589431762695, "learning_rate": 3.577787197965239e-07, "loss": 0.548, "mean_token_accuracy": 0.8240444660186768, "num_tokens": 32142031.0, "step": 845 }, { "epoch": 0.10761989568757156, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.3104987144470215, "learning_rate": 3.5820262823230177e-07, "loss": 0.4802, "mean_token_accuracy": 0.848953127861023, "num_tokens": 32182686.0, "step": 846 }, { "epoch": 0.10774710596616206, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.464416027069092, "learning_rate": 3.5862653666807967e-07, "loss": 0.5754, "mean_token_accuracy": 0.8168810606002808, "num_tokens": 32218027.0, "step": 847 }, { "epoch": 0.10787431624475258, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.399770736694336, "learning_rate": 3.5905044510385757e-07, "loss": 0.5305, "mean_token_accuracy": 0.8335256576538086, "num_tokens": 32257854.0, "step": 848 }, { "epoch": 0.10800152652334309, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.335811138153076, "learning_rate": 3.594743535396354e-07, "loss": 0.5137, "mean_token_accuracy": 0.836128830909729, "num_tokens": 32295019.0, "step": 849 }, { "epoch": 0.10812873680193359, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 7.361173629760742e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.373592376708984, "learning_rate": 3.5989826197541326e-07, "loss": 0.5629, "mean_token_accuracy": 0.8204282522201538, "num_tokens": 32335652.0, "step": 850 }, { "epoch": 0.1082559470805241, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 7.361173629760742e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.361814498901367, "learning_rate": 3.6032217041119116e-07, "loss": 0.5108, "mean_token_accuracy": 0.8325618505477905, "num_tokens": 32377249.0, "step": 851 }, { "epoch": 0.10838315735911462, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 7.361173629760742e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.2824530601501465, "learning_rate": 3.6074607884696906e-07, "loss": 0.4814, "mean_token_accuracy": 0.8468892574310303, "num_tokens": 32422467.0, "step": 852 }, { "epoch": 0.10851036763770512, "ewc_loss": 0.0036773681640625, "ewc_loss_diag": 7.361173629760742e-06, "ewc_loss_parallel": 2.950429916381836e-06, "grad_norm": 5.344604969024658, "learning_rate": 3.611699872827469e-07, "loss": 0.4719, "mean_token_accuracy": 0.8454257249832153, "num_tokens": 32462120.0, "step": 853 }, { "epoch": 0.10863757791629564, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.475795269012451, "learning_rate": 3.6159389571852475e-07, "loss": 0.5919, "mean_token_accuracy": 0.8133776187896729, "num_tokens": 32498484.0, "step": 854 }, { "epoch": 0.10876478819488615, "ewc_loss": 0.0037689208984375, "ewc_loss_diag": 7.420778274536133e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 5.3650946617126465, "learning_rate": 3.6201780415430265e-07, "loss": 0.5281, "mean_token_accuracy": 0.8314574360847473, "num_tokens": 32541536.0, "step": 855 }, { "epoch": 0.10889199847347665, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.420778274536133e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.383457183837891, "learning_rate": 3.6244171259008055e-07, "loss": 0.551, "mean_token_accuracy": 0.8281469345092773, "num_tokens": 32579750.0, "step": 856 }, { "epoch": 0.10901920875206716, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.404189586639404, "learning_rate": 3.628656210258584e-07, "loss": 0.5588, "mean_token_accuracy": 0.8205921053886414, "num_tokens": 32617518.0, "step": 857 }, { "epoch": 0.10914641903065768, "ewc_loss": 0.003753662109375, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 5.375154495239258, "learning_rate": 3.6328952946163624e-07, "loss": 0.5004, "mean_token_accuracy": 0.8393217325210571, "num_tokens": 32653931.0, "step": 858 }, { "epoch": 0.1092736293092482, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.339995861053467, "learning_rate": 3.6371343789741414e-07, "loss": 0.4886, "mean_token_accuracy": 0.8427141308784485, "num_tokens": 32693797.0, "step": 859 }, { "epoch": 0.1094008395878387, "ewc_loss": 0.003753662109375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.381677627563477, "learning_rate": 3.6413734633319204e-07, "loss": 0.5365, "mean_token_accuracy": 0.8299547433853149, "num_tokens": 32732432.0, "step": 860 }, { "epoch": 0.10952804986642921, "ewc_loss": 0.003753662109375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.426253795623779, "learning_rate": 3.645612547689699e-07, "loss": 0.4898, "mean_token_accuracy": 0.8381227254867554, "num_tokens": 32769225.0, "step": 861 }, { "epoch": 0.10965526014501972, "ewc_loss": 0.0037689208984375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 5.353043556213379, "learning_rate": 3.6498516320474773e-07, "loss": 0.4831, "mean_token_accuracy": 0.84555584192276, "num_tokens": 32812027.0, "step": 862 }, { "epoch": 0.10978247042361022, "ewc_loss": 0.0037689208984375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 5.36041784286499, "learning_rate": 3.6540907164052563e-07, "loss": 0.4626, "mean_token_accuracy": 0.8516485095024109, "num_tokens": 32849334.0, "step": 863 }, { "epoch": 0.10990968070220074, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.440891265869141, "learning_rate": 3.658329800763035e-07, "loss": 0.5561, "mean_token_accuracy": 0.8210874199867249, "num_tokens": 32888422.0, "step": 864 }, { "epoch": 0.11003689098079125, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.417800426483154, "learning_rate": 3.662568885120814e-07, "loss": 0.5157, "mean_token_accuracy": 0.8338244557380676, "num_tokens": 32922672.0, "step": 865 }, { "epoch": 0.11016410125938175, "ewc_loss": 0.003753662109375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.444011688232422, "learning_rate": 3.666807969478592e-07, "loss": 0.5455, "mean_token_accuracy": 0.8231269121170044, "num_tokens": 32961694.0, "step": 866 }, { "epoch": 0.11029131153797227, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.4431281089782715, "learning_rate": 3.671047053836371e-07, "loss": 0.5346, "mean_token_accuracy": 0.8269819617271423, "num_tokens": 32998595.0, "step": 867 }, { "epoch": 0.11041852181656278, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.346572399139404, "learning_rate": 3.6752861381941497e-07, "loss": 0.4946, "mean_token_accuracy": 0.8398584723472595, "num_tokens": 33035884.0, "step": 868 }, { "epoch": 0.11054573209515328, "ewc_loss": 0.003753662109375, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.453129291534424, "learning_rate": 3.6795252225519287e-07, "loss": 0.5561, "mean_token_accuracy": 0.8246239423751831, "num_tokens": 33073335.0, "step": 869 }, { "epoch": 0.1106729423737438, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.539987564086914e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 5.357392311096191, "learning_rate": 3.6837643069097077e-07, "loss": 0.4561, "mean_token_accuracy": 0.8513153195381165, "num_tokens": 33111443.0, "step": 870 }, { "epoch": 0.11080015265233431, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.539987564086914e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 5.386466979980469, "learning_rate": 3.688003391267486e-07, "loss": 0.5158, "mean_token_accuracy": 0.8372331261634827, "num_tokens": 33156362.0, "step": 871 }, { "epoch": 0.11092736293092482, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.539987564086914e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.368136405944824, "learning_rate": 3.6922424756252646e-07, "loss": 0.4845, "mean_token_accuracy": 0.8434672355651855, "num_tokens": 33204968.0, "step": 872 }, { "epoch": 0.11105457320951533, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.539987564086914e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.43428373336792, "learning_rate": 3.6964815599830436e-07, "loss": 0.5317, "mean_token_accuracy": 0.8269811272621155, "num_tokens": 33239265.0, "step": 873 }, { "epoch": 0.11118178348810584, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.512871742248535, "learning_rate": 3.7007206443408226e-07, "loss": 0.4652, "mean_token_accuracy": 0.8508101105690002, "num_tokens": 33275897.0, "step": 874 }, { "epoch": 0.11130899376669635, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.4743194580078125, "learning_rate": 3.704959728698601e-07, "loss": 0.5877, "mean_token_accuracy": 0.8138308525085449, "num_tokens": 33313966.0, "step": 875 }, { "epoch": 0.11143620404528685, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.457089900970459, "learning_rate": 3.7091988130563795e-07, "loss": 0.5333, "mean_token_accuracy": 0.830181896686554, "num_tokens": 33352909.0, "step": 876 }, { "epoch": 0.11156341432387737, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.377175807952881, "learning_rate": 3.7134378974141585e-07, "loss": 0.5161, "mean_token_accuracy": 0.8363171815872192, "num_tokens": 33400120.0, "step": 877 }, { "epoch": 0.11169062460246788, "ewc_loss": 0.0037994384765625, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.491907596588135, "learning_rate": 3.7176769817719375e-07, "loss": 0.5432, "mean_token_accuracy": 0.8267723321914673, "num_tokens": 33436703.0, "step": 878 }, { "epoch": 0.11181783488105838, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.442736625671387, "learning_rate": 3.7219160661297154e-07, "loss": 0.4526, "mean_token_accuracy": 0.8548046350479126, "num_tokens": 33473515.0, "step": 879 }, { "epoch": 0.1119450451596489, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 5.445772171020508, "learning_rate": 3.7261551504874944e-07, "loss": 0.5143, "mean_token_accuracy": 0.8373129367828369, "num_tokens": 33507659.0, "step": 880 }, { "epoch": 0.11207225543823941, "ewc_loss": 0.0038299560546875, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.417232036590576, "learning_rate": 3.7303942348452734e-07, "loss": 0.4604, "mean_token_accuracy": 0.8485777378082275, "num_tokens": 33542430.0, "step": 881 }, { "epoch": 0.11219946571682991, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.3633503913879395, "learning_rate": 3.7346333192030524e-07, "loss": 0.4678, "mean_token_accuracy": 0.8523493409156799, "num_tokens": 33582579.0, "step": 882 }, { "epoch": 0.11232667599542043, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.420565128326416, "learning_rate": 3.7388724035608303e-07, "loss": 0.4913, "mean_token_accuracy": 0.8456214666366577, "num_tokens": 33625928.0, "step": 883 }, { "epoch": 0.11245388627401094, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.476117134094238, "learning_rate": 3.7431114879186093e-07, "loss": 0.5326, "mean_token_accuracy": 0.8309074640274048, "num_tokens": 33661325.0, "step": 884 }, { "epoch": 0.11258109655260146, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.447259426116943, "learning_rate": 3.7473505722763883e-07, "loss": 0.5541, "mean_token_accuracy": 0.8233067393302917, "num_tokens": 33696472.0, "step": 885 }, { "epoch": 0.11270830683119196, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.450807094573975, "learning_rate": 3.7515896566341673e-07, "loss": 0.509, "mean_token_accuracy": 0.8344478607177734, "num_tokens": 33732637.0, "step": 886 }, { "epoch": 0.11283551710978247, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.394861221313477, "learning_rate": 3.755828740991945e-07, "loss": 0.5331, "mean_token_accuracy": 0.8298227190971375, "num_tokens": 33775106.0, "step": 887 }, { "epoch": 0.11296272738837299, "ewc_loss": 0.0038299560546875, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.464323043823242, "learning_rate": 3.760067825349724e-07, "loss": 0.4957, "mean_token_accuracy": 0.8418998718261719, "num_tokens": 33811297.0, "step": 888 }, { "epoch": 0.11308993766696349, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.377865791320801, "learning_rate": 3.764306909707503e-07, "loss": 0.4577, "mean_token_accuracy": 0.8510748744010925, "num_tokens": 33854038.0, "step": 889 }, { "epoch": 0.113217147945554, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.53415060043335, "learning_rate": 3.768545994065282e-07, "loss": 0.5138, "mean_token_accuracy": 0.8339253067970276, "num_tokens": 33884929.0, "step": 890 }, { "epoch": 0.11334435822414451, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.469822406768799, "learning_rate": 3.77278507842306e-07, "loss": 0.5222, "mean_token_accuracy": 0.8346359729766846, "num_tokens": 33926065.0, "step": 891 }, { "epoch": 0.11347156850273502, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.416670322418213, "learning_rate": 3.777024162780839e-07, "loss": 0.4729, "mean_token_accuracy": 0.8447637557983398, "num_tokens": 33965557.0, "step": 892 }, { "epoch": 0.11359877878132553, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.452437877655029, "learning_rate": 3.781263247138618e-07, "loss": 0.4818, "mean_token_accuracy": 0.8436317443847656, "num_tokens": 34004132.0, "step": 893 }, { "epoch": 0.11372598905991604, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.502183437347412, "learning_rate": 3.785502331496397e-07, "loss": 0.5093, "mean_token_accuracy": 0.8295576572418213, "num_tokens": 34039781.0, "step": 894 }, { "epoch": 0.11385319933850654, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.490232467651367, "learning_rate": 3.789741415854175e-07, "loss": 0.5352, "mean_token_accuracy": 0.8276470899581909, "num_tokens": 34080487.0, "step": 895 }, { "epoch": 0.11398040961709706, "ewc_loss": 0.0038909912109375, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.453911781311035, "learning_rate": 3.793980500211954e-07, "loss": 0.5123, "mean_token_accuracy": 0.835314929485321, "num_tokens": 34119913.0, "step": 896 }, { "epoch": 0.11410761989568757, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.486997604370117, "learning_rate": 3.798219584569733e-07, "loss": 0.4467, "mean_token_accuracy": 0.8546782732009888, "num_tokens": 34155811.0, "step": 897 }, { "epoch": 0.11423483017427809, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.517231464385986, "learning_rate": 3.8024586689275115e-07, "loss": 0.5682, "mean_token_accuracy": 0.8230823278427124, "num_tokens": 34195022.0, "step": 898 }, { "epoch": 0.11436204045286859, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.481856822967529, "learning_rate": 3.80669775328529e-07, "loss": 0.5007, "mean_token_accuracy": 0.8356601595878601, "num_tokens": 34229182.0, "step": 899 }, { "epoch": 0.1144892507314591, "ewc_loss": 0.0038909912109375, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.451247692108154, "learning_rate": 3.810936837643069e-07, "loss": 0.462, "mean_token_accuracy": 0.8490463495254517, "num_tokens": 34266931.0, "step": 900 }, { "epoch": 0.11461646101004962, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 5.489643573760986, "learning_rate": 3.815175922000848e-07, "loss": 0.4545, "mean_token_accuracy": 0.8493484854698181, "num_tokens": 34301705.0, "step": 901 }, { "epoch": 0.11474367128864012, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.48473596572876, "learning_rate": 3.8194150063586264e-07, "loss": 0.5372, "mean_token_accuracy": 0.8328169584274292, "num_tokens": 34340975.0, "step": 902 }, { "epoch": 0.11487088156723063, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.495301723480225, "learning_rate": 3.823654090716405e-07, "loss": 0.5049, "mean_token_accuracy": 0.8357441425323486, "num_tokens": 34380024.0, "step": 903 }, { "epoch": 0.11499809184582115, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.487751483917236, "learning_rate": 3.827893175074184e-07, "loss": 0.5264, "mean_token_accuracy": 0.838287353515625, "num_tokens": 34419654.0, "step": 904 }, { "epoch": 0.11512530212441165, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1888484954833984e-06, "grad_norm": 5.523016452789307, "learning_rate": 3.832132259431963e-07, "loss": 0.4571, "mean_token_accuracy": 0.8534663319587708, "num_tokens": 34453034.0, "step": 905 }, { "epoch": 0.11525251240300216, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 5.495937347412109, "learning_rate": 3.8363713437897413e-07, "loss": 0.5626, "mean_token_accuracy": 0.819601833820343, "num_tokens": 34493694.0, "step": 906 }, { "epoch": 0.11537972268159268, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.5250396728515625, "learning_rate": 3.8406104281475197e-07, "loss": 0.5544, "mean_token_accuracy": 0.825014591217041, "num_tokens": 34531636.0, "step": 907 }, { "epoch": 0.11550693296018319, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.5288543701171875, "learning_rate": 3.8448495125052987e-07, "loss": 0.5392, "mean_token_accuracy": 0.8258531093597412, "num_tokens": 34570851.0, "step": 908 }, { "epoch": 0.11563414323877369, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.475500106811523, "learning_rate": 3.8490885968630777e-07, "loss": 0.491, "mean_token_accuracy": 0.8421214818954468, "num_tokens": 34613314.0, "step": 909 }, { "epoch": 0.1157613535173642, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.469806671142578, "learning_rate": 3.853327681220856e-07, "loss": 0.5066, "mean_token_accuracy": 0.835711658000946, "num_tokens": 34656701.0, "step": 910 }, { "epoch": 0.11588856379595472, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.5405378341674805, "learning_rate": 3.8575667655786346e-07, "loss": 0.4818, "mean_token_accuracy": 0.8459020853042603, "num_tokens": 34688595.0, "step": 911 }, { "epoch": 0.11601577407454522, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.472301959991455, "learning_rate": 3.8618058499364136e-07, "loss": 0.5002, "mean_token_accuracy": 0.8401560187339783, "num_tokens": 34726609.0, "step": 912 }, { "epoch": 0.11614298435313573, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2186508178710938e-06, "grad_norm": 5.553167343139648, "learning_rate": 3.8660449342941926e-07, "loss": 0.5565, "mean_token_accuracy": 0.8253828287124634, "num_tokens": 34769213.0, "step": 913 }, { "epoch": 0.11627019463172625, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.545374870300293, "learning_rate": 3.870284018651971e-07, "loss": 0.5294, "mean_token_accuracy": 0.8316512107849121, "num_tokens": 34806722.0, "step": 914 }, { "epoch": 0.11639740491031675, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2186508178710938e-06, "grad_norm": 5.522469520568848, "learning_rate": 3.8745231030097495e-07, "loss": 0.4853, "mean_token_accuracy": 0.8461107015609741, "num_tokens": 34846408.0, "step": 915 }, { "epoch": 0.11652461518890726, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.538743019104004, "learning_rate": 3.8787621873675285e-07, "loss": 0.4728, "mean_token_accuracy": 0.8465108275413513, "num_tokens": 34883791.0, "step": 916 }, { "epoch": 0.11665182546749778, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.512451171875, "learning_rate": 3.883001271725307e-07, "loss": 0.4685, "mean_token_accuracy": 0.8483405709266663, "num_tokens": 34922768.0, "step": 917 }, { "epoch": 0.11677903574608828, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.518549919128418, "learning_rate": 3.887240356083086e-07, "loss": 0.4481, "mean_token_accuracy": 0.8536052703857422, "num_tokens": 34962184.0, "step": 918 }, { "epoch": 0.11690624602467879, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2186508178710938e-06, "grad_norm": 5.712454795837402, "learning_rate": 3.8914794404408644e-07, "loss": 0.4925, "mean_token_accuracy": 0.8398605585098267, "num_tokens": 34991455.0, "step": 919 }, { "epoch": 0.1170334563032693, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.621676445007324, "learning_rate": 3.8957185247986434e-07, "loss": 0.5219, "mean_token_accuracy": 0.8307433128356934, "num_tokens": 35025044.0, "step": 920 }, { "epoch": 0.11716066658185982, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2335519790649414e-06, "grad_norm": 5.516578197479248, "learning_rate": 3.899957609156422e-07, "loss": 0.495, "mean_token_accuracy": 0.8396027088165283, "num_tokens": 35064564.0, "step": 921 }, { "epoch": 0.11728787686045032, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.480911731719971, "learning_rate": 3.904196693514201e-07, "loss": 0.5122, "mean_token_accuracy": 0.8366198539733887, "num_tokens": 35109786.0, "step": 922 }, { "epoch": 0.11741508713904084, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2335519790649414e-06, "grad_norm": 5.571168422698975, "learning_rate": 3.9084357778719793e-07, "loss": 0.5752, "mean_token_accuracy": 0.8170187473297119, "num_tokens": 35148351.0, "step": 923 }, { "epoch": 0.11754229741763135, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 5.577431678771973, "learning_rate": 3.9126748622297583e-07, "loss": 0.5392, "mean_token_accuracy": 0.8292286396026611, "num_tokens": 35183464.0, "step": 924 }, { "epoch": 0.11766950769622185, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.248453140258789e-06, "grad_norm": 5.5547943115234375, "learning_rate": 3.916913946587537e-07, "loss": 0.5216, "mean_token_accuracy": 0.8302924633026123, "num_tokens": 35218561.0, "step": 925 }, { "epoch": 0.11779671797481236, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.52550745010376, "learning_rate": 3.921153030945316e-07, "loss": 0.5365, "mean_token_accuracy": 0.827275276184082, "num_tokens": 35257740.0, "step": 926 }, { "epoch": 0.11792392825340288, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.617215633392334, "learning_rate": 3.925392115303094e-07, "loss": 0.493, "mean_token_accuracy": 0.8410113453865051, "num_tokens": 35291632.0, "step": 927 }, { "epoch": 0.11805113853199338, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3080577850341797e-06, "grad_norm": 5.596765041351318, "learning_rate": 3.929631199660873e-07, "loss": 0.4969, "mean_token_accuracy": 0.8370582461357117, "num_tokens": 35327686.0, "step": 928 }, { "epoch": 0.1181783488105839, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.529617786407471, "learning_rate": 3.9338702840186517e-07, "loss": 0.4981, "mean_token_accuracy": 0.8368269205093384, "num_tokens": 35364480.0, "step": 929 }, { "epoch": 0.11830555908917441, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 5.4961676597595215, "learning_rate": 3.9381093683764307e-07, "loss": 0.5039, "mean_token_accuracy": 0.8383885622024536, "num_tokens": 35406017.0, "step": 930 }, { "epoch": 0.11843276936776491, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.5139031410217285, "learning_rate": 3.942348452734209e-07, "loss": 0.5629, "mean_token_accuracy": 0.8198274374008179, "num_tokens": 35448447.0, "step": 931 }, { "epoch": 0.11855997964635542, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.530797958374023, "learning_rate": 3.946587537091988e-07, "loss": 0.5116, "mean_token_accuracy": 0.8335070610046387, "num_tokens": 35488120.0, "step": 932 }, { "epoch": 0.11868718992494594, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.682675361633301, "learning_rate": 3.9508266214497666e-07, "loss": 0.486, "mean_token_accuracy": 0.8426061868667603, "num_tokens": 35531759.0, "step": 933 }, { "epoch": 0.11881440020353645, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.548305034637451, "learning_rate": 3.9550657058075456e-07, "loss": 0.4699, "mean_token_accuracy": 0.8500513434410095, "num_tokens": 35571476.0, "step": 934 }, { "epoch": 0.11894161048212695, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3080577850341797e-06, "grad_norm": 5.545218467712402, "learning_rate": 3.959304790165324e-07, "loss": 0.5108, "mean_token_accuracy": 0.837592363357544, "num_tokens": 35608440.0, "step": 935 }, { "epoch": 0.11906882076071747, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.644066333770752, "learning_rate": 3.9635438745231025e-07, "loss": 0.5058, "mean_token_accuracy": 0.8386280536651611, "num_tokens": 35641861.0, "step": 936 }, { "epoch": 0.11919603103930798, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.594743251800537, "learning_rate": 3.9677829588808815e-07, "loss": 0.5275, "mean_token_accuracy": 0.8299267888069153, "num_tokens": 35684626.0, "step": 937 }, { "epoch": 0.11932324131789848, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.576115131378174, "learning_rate": 3.9720220432386605e-07, "loss": 0.534, "mean_token_accuracy": 0.8278307914733887, "num_tokens": 35718125.0, "step": 938 }, { "epoch": 0.119450451596489, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.600037574768066, "learning_rate": 3.976261127596439e-07, "loss": 0.4777, "mean_token_accuracy": 0.8473588228225708, "num_tokens": 35752818.0, "step": 939 }, { "epoch": 0.11957766187507951, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.551910400390625, "learning_rate": 3.9805002119542174e-07, "loss": 0.458, "mean_token_accuracy": 0.851345419883728, "num_tokens": 35787097.0, "step": 940 }, { "epoch": 0.11970487215367001, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 5.566970348358154, "learning_rate": 3.9847392963119964e-07, "loss": 0.5548, "mean_token_accuracy": 0.8260139226913452, "num_tokens": 35825887.0, "step": 941 }, { "epoch": 0.11983208243226053, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 5.5686469078063965, "learning_rate": 3.9889783806697754e-07, "loss": 0.4835, "mean_token_accuracy": 0.8461618423461914, "num_tokens": 35863592.0, "step": 942 }, { "epoch": 0.11995929271085104, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.6066436767578125, "learning_rate": 3.993217465027554e-07, "loss": 0.5583, "mean_token_accuracy": 0.8253811597824097, "num_tokens": 35904483.0, "step": 943 }, { "epoch": 0.12008650298944154, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.608839511871338, "learning_rate": 3.9974565493853323e-07, "loss": 0.4457, "mean_token_accuracy": 0.8561708927154541, "num_tokens": 35938662.0, "step": 944 }, { "epoch": 0.12021371326803205, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 5.590915679931641, "learning_rate": 4.0016956337431113e-07, "loss": 0.5097, "mean_token_accuracy": 0.8353065848350525, "num_tokens": 35975176.0, "step": 945 }, { "epoch": 0.12034092354662257, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 5.596612930297852, "learning_rate": 4.0059347181008903e-07, "loss": 0.4957, "mean_token_accuracy": 0.839326798915863, "num_tokens": 36008893.0, "step": 946 }, { "epoch": 0.12046813382521308, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.532066822052002, "learning_rate": 4.010173802458669e-07, "loss": 0.4553, "mean_token_accuracy": 0.8537365198135376, "num_tokens": 36049882.0, "step": 947 }, { "epoch": 0.12059534410380358, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 5.564456462860107, "learning_rate": 4.014412886816447e-07, "loss": 0.5688, "mean_token_accuracy": 0.8220798969268799, "num_tokens": 36090615.0, "step": 948 }, { "epoch": 0.1207225543823941, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.614267349243164, "learning_rate": 4.018651971174226e-07, "loss": 0.499, "mean_token_accuracy": 0.837621808052063, "num_tokens": 36128159.0, "step": 949 }, { "epoch": 0.12084976466098461, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.536113262176514, "learning_rate": 4.022891055532005e-07, "loss": 0.5128, "mean_token_accuracy": 0.8356413245201111, "num_tokens": 36172851.0, "step": 950 }, { "epoch": 0.12097697493957511, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 5.62891960144043, "learning_rate": 4.0271301398897837e-07, "loss": 0.5632, "mean_token_accuracy": 0.8239392042160034, "num_tokens": 36214282.0, "step": 951 }, { "epoch": 0.12110418521816563, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.641170978546143, "learning_rate": 4.031369224247562e-07, "loss": 0.4848, "mean_token_accuracy": 0.8391770720481873, "num_tokens": 36248619.0, "step": 952 }, { "epoch": 0.12123139549675614, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.571138381958008, "learning_rate": 4.035608308605341e-07, "loss": 0.5208, "mean_token_accuracy": 0.8314319849014282, "num_tokens": 36289070.0, "step": 953 }, { "epoch": 0.12135860577534664, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.641526699066162, "learning_rate": 4.03984739296312e-07, "loss": 0.5206, "mean_token_accuracy": 0.8353505730628967, "num_tokens": 36330883.0, "step": 954 }, { "epoch": 0.12148581605393716, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.582963466644287, "learning_rate": 4.044086477320898e-07, "loss": 0.5006, "mean_token_accuracy": 0.8392833471298218, "num_tokens": 36371507.0, "step": 955 }, { "epoch": 0.12161302633252767, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.632136821746826, "learning_rate": 4.048325561678677e-07, "loss": 0.5233, "mean_token_accuracy": 0.8315554261207581, "num_tokens": 36407262.0, "step": 956 }, { "epoch": 0.12174023661111817, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.625185489654541, "learning_rate": 4.052564646036456e-07, "loss": 0.5194, "mean_token_accuracy": 0.8346617221832275, "num_tokens": 36448089.0, "step": 957 }, { "epoch": 0.12186744688970869, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.616549491882324, "learning_rate": 4.056803730394235e-07, "loss": 0.5319, "mean_token_accuracy": 0.8275108337402344, "num_tokens": 36484022.0, "step": 958 }, { "epoch": 0.1219946571682992, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.654906272888184, "learning_rate": 4.061042814752013e-07, "loss": 0.5083, "mean_token_accuracy": 0.8378401398658752, "num_tokens": 36519385.0, "step": 959 }, { "epoch": 0.12212186744688971, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.734455585479736, "learning_rate": 4.065281899109792e-07, "loss": 0.5117, "mean_token_accuracy": 0.834784984588623, "num_tokens": 36551943.0, "step": 960 }, { "epoch": 0.12224907772548022, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.632850646972656, "learning_rate": 4.069520983467571e-07, "loss": 0.4586, "mean_token_accuracy": 0.852583646774292, "num_tokens": 36590927.0, "step": 961 }, { "epoch": 0.12237628800407073, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.649107933044434, "learning_rate": 4.07376006782535e-07, "loss": 0.4905, "mean_token_accuracy": 0.8399733304977417, "num_tokens": 36625199.0, "step": 962 }, { "epoch": 0.12250349828266124, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.660272121429443, "learning_rate": 4.077999152183128e-07, "loss": 0.453, "mean_token_accuracy": 0.8526328802108765, "num_tokens": 36658981.0, "step": 963 }, { "epoch": 0.12263070856125174, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.580744743347168, "learning_rate": 4.082238236540907e-07, "loss": 0.5136, "mean_token_accuracy": 0.8377809524536133, "num_tokens": 36704214.0, "step": 964 }, { "epoch": 0.12275791883984226, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.638283729553223, "learning_rate": 4.086477320898686e-07, "loss": 0.4831, "mean_token_accuracy": 0.8423789739608765, "num_tokens": 36744782.0, "step": 965 }, { "epoch": 0.12288512911843277, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.616425037384033, "learning_rate": 4.090716405256465e-07, "loss": 0.5178, "mean_token_accuracy": 0.8321523666381836, "num_tokens": 36783645.0, "step": 966 }, { "epoch": 0.12301233939702327, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 3.427267074584961e-06, "grad_norm": 5.710367679595947, "learning_rate": 4.094955489614243e-07, "loss": 0.486, "mean_token_accuracy": 0.8414427638053894, "num_tokens": 36817539.0, "step": 967 }, { "epoch": 0.12313954967561379, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.684959411621094, "learning_rate": 4.099194573972022e-07, "loss": 0.4879, "mean_token_accuracy": 0.8413113355636597, "num_tokens": 36859144.0, "step": 968 }, { "epoch": 0.1232667599542043, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.623097896575928, "learning_rate": 4.1034336583298007e-07, "loss": 0.5307, "mean_token_accuracy": 0.8296185731887817, "num_tokens": 36900833.0, "step": 969 }, { "epoch": 0.1233939702327948, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.634173393249512, "learning_rate": 4.1076727426875797e-07, "loss": 0.5197, "mean_token_accuracy": 0.832720160484314, "num_tokens": 36940935.0, "step": 970 }, { "epoch": 0.12352118051138532, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.721611499786377, "learning_rate": 4.1119118270453577e-07, "loss": 0.5812, "mean_token_accuracy": 0.8137099742889404, "num_tokens": 36979874.0, "step": 971 }, { "epoch": 0.12364839078997583, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.637701034545898, "learning_rate": 4.1161509114031366e-07, "loss": 0.4634, "mean_token_accuracy": 0.8516312837600708, "num_tokens": 37017917.0, "step": 972 }, { "epoch": 0.12377560106856635, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.644421577453613, "learning_rate": 4.1203899957609156e-07, "loss": 0.5277, "mean_token_accuracy": 0.8313487768173218, "num_tokens": 37054914.0, "step": 973 }, { "epoch": 0.12390281134715685, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.427267074584961e-06, "grad_norm": 5.710068702697754, "learning_rate": 4.124629080118694e-07, "loss": 0.4778, "mean_token_accuracy": 0.8425791263580322, "num_tokens": 37088277.0, "step": 974 }, { "epoch": 0.12403002162574736, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.605405330657959, "learning_rate": 4.1288681644764726e-07, "loss": 0.5236, "mean_token_accuracy": 0.830846905708313, "num_tokens": 37126963.0, "step": 975 }, { "epoch": 0.12415723190433788, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.675325870513916, "learning_rate": 4.1331072488342515e-07, "loss": 0.5365, "mean_token_accuracy": 0.8288889527320862, "num_tokens": 37161948.0, "step": 976 }, { "epoch": 0.12428444218292838, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 5.6701812744140625, "learning_rate": 4.1373463331920305e-07, "loss": 0.4518, "mean_token_accuracy": 0.853705644607544, "num_tokens": 37198552.0, "step": 977 }, { "epoch": 0.12441165246151889, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.664369106292725, "learning_rate": 4.141585417549809e-07, "loss": 0.5216, "mean_token_accuracy": 0.8310599327087402, "num_tokens": 37235990.0, "step": 978 }, { "epoch": 0.1245388627401094, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.7126078605651855, "learning_rate": 4.1458245019075875e-07, "loss": 0.4811, "mean_token_accuracy": 0.8434969186782837, "num_tokens": 37269664.0, "step": 979 }, { "epoch": 0.1246660730186999, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 3.5017728805541992e-06, "grad_norm": 5.641238212585449, "learning_rate": 4.1500635862653664e-07, "loss": 0.5125, "mean_token_accuracy": 0.836514949798584, "num_tokens": 37308647.0, "step": 980 }, { "epoch": 0.12479328329729042, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 5.642500400543213, "learning_rate": 4.1543026706231454e-07, "loss": 0.4619, "mean_token_accuracy": 0.8509602546691895, "num_tokens": 37345953.0, "step": 981 }, { "epoch": 0.12492049357588093, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.678432941436768, "learning_rate": 4.158541754980924e-07, "loss": 0.4579, "mean_token_accuracy": 0.8536797165870667, "num_tokens": 37382735.0, "step": 982 }, { "epoch": 0.12504770385447145, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.667634010314941, "learning_rate": 4.1627808393387024e-07, "loss": 0.5242, "mean_token_accuracy": 0.8284575939178467, "num_tokens": 37419605.0, "step": 983 }, { "epoch": 0.12517491413306195, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 5.760161399841309, "learning_rate": 4.1670199236964813e-07, "loss": 0.5164, "mean_token_accuracy": 0.829373836517334, "num_tokens": 37455395.0, "step": 984 }, { "epoch": 0.12530212441165245, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.5017728805541992e-06, "grad_norm": 5.681478500366211, "learning_rate": 4.1712590080542603e-07, "loss": 0.4699, "mean_token_accuracy": 0.8479921817779541, "num_tokens": 37492208.0, "step": 985 }, { "epoch": 0.12542933469024298, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.650415420532227, "learning_rate": 4.175498092412039e-07, "loss": 0.5149, "mean_token_accuracy": 0.8369159698486328, "num_tokens": 37533675.0, "step": 986 }, { "epoch": 0.12555654496883348, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.5017728805541992e-06, "grad_norm": 5.722732067108154, "learning_rate": 4.179737176769817e-07, "loss": 0.4794, "mean_token_accuracy": 0.845920205116272, "num_tokens": 37572792.0, "step": 987 }, { "epoch": 0.12568375524742398, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.693714141845703, "learning_rate": 4.183976261127596e-07, "loss": 0.4763, "mean_token_accuracy": 0.8442909717559814, "num_tokens": 37611481.0, "step": 988 }, { "epoch": 0.1258109655260145, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.734557151794434, "learning_rate": 4.1882153454853747e-07, "loss": 0.5547, "mean_token_accuracy": 0.8221403360366821, "num_tokens": 37648253.0, "step": 989 }, { "epoch": 0.125938175804605, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.693888187408447, "learning_rate": 4.1924544298431537e-07, "loss": 0.4883, "mean_token_accuracy": 0.8435088396072388, "num_tokens": 37687086.0, "step": 990 }, { "epoch": 0.12606538608319554, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.662968158721924, "learning_rate": 4.196693514200932e-07, "loss": 0.4823, "mean_token_accuracy": 0.8442673087120056, "num_tokens": 37730928.0, "step": 991 }, { "epoch": 0.12619259636178604, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.684433937072754, "learning_rate": 4.200932598558711e-07, "loss": 0.5065, "mean_token_accuracy": 0.8346033096313477, "num_tokens": 37771857.0, "step": 992 }, { "epoch": 0.12631980664037654, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.691900730133057, "learning_rate": 4.2051716829164896e-07, "loss": 0.5097, "mean_token_accuracy": 0.8384286165237427, "num_tokens": 37817517.0, "step": 993 }, { "epoch": 0.12644701691896706, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.732064247131348, "learning_rate": 4.2094107672742686e-07, "loss": 0.5524, "mean_token_accuracy": 0.8282358646392822, "num_tokens": 37855891.0, "step": 994 }, { "epoch": 0.12657422719755757, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.766815662384033, "learning_rate": 4.2136498516320476e-07, "loss": 0.4722, "mean_token_accuracy": 0.849066972732544, "num_tokens": 37888719.0, "step": 995 }, { "epoch": 0.12670143747614807, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.678077220916748, "learning_rate": 4.217888935989826e-07, "loss": 0.524, "mean_token_accuracy": 0.8310878276824951, "num_tokens": 37932968.0, "step": 996 }, { "epoch": 0.1268286477547386, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.7352294921875, "learning_rate": 4.2221280203476045e-07, "loss": 0.5161, "mean_token_accuracy": 0.8323121070861816, "num_tokens": 37974297.0, "step": 997 }, { "epoch": 0.1269558580333291, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 5.713921070098877, "learning_rate": 4.2263671047053835e-07, "loss": 0.5324, "mean_token_accuracy": 0.8297569751739502, "num_tokens": 38020294.0, "step": 998 }, { "epoch": 0.1270830683119196, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.685734272003174, "learning_rate": 4.2306061890631625e-07, "loss": 0.5123, "mean_token_accuracy": 0.8359876871109009, "num_tokens": 38063322.0, "step": 999 }, { "epoch": 0.12721027859051012, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.773543834686279, "learning_rate": 4.234845273420941e-07, "loss": 0.6001, "mean_token_accuracy": 0.8164461851119995, "num_tokens": 38101052.0, "step": 1000 }, { "epoch": 0.12733748886910062, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 3.591179847717285e-06, "grad_norm": 5.698643684387207, "learning_rate": 4.2390843577787194e-07, "loss": 0.5204, "mean_token_accuracy": 0.8325344324111938, "num_tokens": 38145472.0, "step": 1001 }, { "epoch": 0.12746469914769112, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.727570533752441, "learning_rate": 4.2433234421364984e-07, "loss": 0.4974, "mean_token_accuracy": 0.8402469158172607, "num_tokens": 38182662.0, "step": 1002 }, { "epoch": 0.12759190942628165, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 5.740821361541748, "learning_rate": 4.2475625264942774e-07, "loss": 0.4836, "mean_token_accuracy": 0.8455272912979126, "num_tokens": 38224735.0, "step": 1003 }, { "epoch": 0.12771911970487215, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.87535285949707, "learning_rate": 4.251801610852056e-07, "loss": 0.527, "mean_token_accuracy": 0.8307026624679565, "num_tokens": 38255847.0, "step": 1004 }, { "epoch": 0.12784632998346265, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.731954574584961, "learning_rate": 4.2560406952098343e-07, "loss": 0.5403, "mean_token_accuracy": 0.828934907913208, "num_tokens": 38296921.0, "step": 1005 }, { "epoch": 0.12797354026205318, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 5.716384410858154, "learning_rate": 4.2602797795676133e-07, "loss": 0.5503, "mean_token_accuracy": 0.8253089785575867, "num_tokens": 38336408.0, "step": 1006 }, { "epoch": 0.12810075054064368, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.591179847717285e-06, "grad_norm": 5.73333215713501, "learning_rate": 4.2645188639253923e-07, "loss": 0.4602, "mean_token_accuracy": 0.8494685888290405, "num_tokens": 38372809.0, "step": 1007 }, { "epoch": 0.12822796081923418, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.776907920837402, "learning_rate": 4.26875794828317e-07, "loss": 0.5186, "mean_token_accuracy": 0.8321793079376221, "num_tokens": 38413312.0, "step": 1008 }, { "epoch": 0.1283551710978247, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.6209821701049805e-06, "grad_norm": 5.715393543243408, "learning_rate": 4.272997032640949e-07, "loss": 0.5383, "mean_token_accuracy": 0.8275001645088196, "num_tokens": 38457710.0, "step": 1009 }, { "epoch": 0.1284823813764152, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.591179847717285e-06, "grad_norm": 5.770443439483643, "learning_rate": 4.277236116998728e-07, "loss": 0.4984, "mean_token_accuracy": 0.8403226733207703, "num_tokens": 38499630.0, "step": 1010 }, { "epoch": 0.1286095916550057, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.635883331298828e-06, "grad_norm": 5.766862869262695, "learning_rate": 4.281475201356507e-07, "loss": 0.5059, "mean_token_accuracy": 0.8388103246688843, "num_tokens": 38538133.0, "step": 1011 }, { "epoch": 0.12873680193359624, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.758150577545166, "learning_rate": 4.285714285714285e-07, "loss": 0.5381, "mean_token_accuracy": 0.8285427093505859, "num_tokens": 38576816.0, "step": 1012 }, { "epoch": 0.12886401221218674, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.764489650726318, "learning_rate": 4.289953370072064e-07, "loss": 0.4776, "mean_token_accuracy": 0.8452595472335815, "num_tokens": 38615270.0, "step": 1013 }, { "epoch": 0.12899122249077727, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.635883331298828e-06, "grad_norm": 5.825449466705322, "learning_rate": 4.294192454429843e-07, "loss": 0.5291, "mean_token_accuracy": 0.8293888568878174, "num_tokens": 38654173.0, "step": 1014 }, { "epoch": 0.12911843276936777, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.635883331298828e-06, "grad_norm": 5.708775997161865, "learning_rate": 4.298431538787622e-07, "loss": 0.4157, "mean_token_accuracy": 0.866484522819519, "num_tokens": 38695758.0, "step": 1015 }, { "epoch": 0.12924564304795827, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.591179847717285e-06, "grad_norm": 5.883216857910156, "learning_rate": 4.3026706231454e-07, "loss": 0.5612, "mean_token_accuracy": 0.8203484416007996, "num_tokens": 38729524.0, "step": 1016 }, { "epoch": 0.1293728533265488, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.831803798675537, "learning_rate": 4.306909707503179e-07, "loss": 0.5083, "mean_token_accuracy": 0.8404123187065125, "num_tokens": 38767500.0, "step": 1017 }, { "epoch": 0.1295000636051393, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.6507844924926758e-06, "grad_norm": 5.833071231842041, "learning_rate": 4.311148791860958e-07, "loss": 0.4958, "mean_token_accuracy": 0.8392045497894287, "num_tokens": 38798683.0, "step": 1018 }, { "epoch": 0.1296272738837298, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.836963176727295, "learning_rate": 4.315387876218737e-07, "loss": 0.5448, "mean_token_accuracy": 0.8277866244316101, "num_tokens": 38833653.0, "step": 1019 }, { "epoch": 0.12975448416232033, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.839484691619873, "learning_rate": 4.319626960576515e-07, "loss": 0.5428, "mean_token_accuracy": 0.8243901133537292, "num_tokens": 38869428.0, "step": 1020 }, { "epoch": 0.12988169444091083, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.7555623054504395, "learning_rate": 4.323866044934294e-07, "loss": 0.4793, "mean_token_accuracy": 0.8411076068878174, "num_tokens": 38909502.0, "step": 1021 }, { "epoch": 0.13000890471950133, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.793359756469727, "learning_rate": 4.328105129292073e-07, "loss": 0.5034, "mean_token_accuracy": 0.8401160836219788, "num_tokens": 38951730.0, "step": 1022 }, { "epoch": 0.13013611499809186, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.802315711975098, "learning_rate": 4.332344213649852e-07, "loss": 0.4808, "mean_token_accuracy": 0.84232097864151, "num_tokens": 38988981.0, "step": 1023 }, { "epoch": 0.13026332527668236, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.807672500610352, "learning_rate": 4.33658329800763e-07, "loss": 0.4904, "mean_token_accuracy": 0.8429350852966309, "num_tokens": 39026289.0, "step": 1024 }, { "epoch": 0.13039053555527286, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.818360328674316, "learning_rate": 4.340822382365409e-07, "loss": 0.4806, "mean_token_accuracy": 0.8433048725128174, "num_tokens": 39069113.0, "step": 1025 }, { "epoch": 0.13051774583386339, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.750469207763672, "learning_rate": 4.345061466723188e-07, "loss": 0.4647, "mean_token_accuracy": 0.8501478433609009, "num_tokens": 39112991.0, "step": 1026 }, { "epoch": 0.13064495611245389, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6507844924926758e-06, "grad_norm": 5.83869743347168, "learning_rate": 4.3493005510809663e-07, "loss": 0.5274, "mean_token_accuracy": 0.8323985934257507, "num_tokens": 39149593.0, "step": 1027 }, { "epoch": 0.1307721663910444, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.816844940185547, "learning_rate": 4.353539635438745e-07, "loss": 0.4914, "mean_token_accuracy": 0.8406307697296143, "num_tokens": 39186570.0, "step": 1028 }, { "epoch": 0.13089937666963491, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.876345634460449, "learning_rate": 4.357778719796524e-07, "loss": 0.5544, "mean_token_accuracy": 0.8203009963035583, "num_tokens": 39220666.0, "step": 1029 }, { "epoch": 0.13102658694822542, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.782683849334717, "learning_rate": 4.362017804154303e-07, "loss": 0.4731, "mean_token_accuracy": 0.8487045764923096, "num_tokens": 39258372.0, "step": 1030 }, { "epoch": 0.13115379722681592, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.8134965896606445, "learning_rate": 4.366256888512081e-07, "loss": 0.4686, "mean_token_accuracy": 0.8495599031448364, "num_tokens": 39293647.0, "step": 1031 }, { "epoch": 0.13128100750540644, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 5.828840255737305, "learning_rate": 4.3704959728698597e-07, "loss": 0.4964, "mean_token_accuracy": 0.8382415771484375, "num_tokens": 39332811.0, "step": 1032 }, { "epoch": 0.13140821778399694, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.849754333496094, "learning_rate": 4.3747350572276386e-07, "loss": 0.4984, "mean_token_accuracy": 0.8370296955108643, "num_tokens": 39365971.0, "step": 1033 }, { "epoch": 0.13153542806258745, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.81999397277832, "learning_rate": 4.3789741415854176e-07, "loss": 0.5099, "mean_token_accuracy": 0.8370739817619324, "num_tokens": 39403178.0, "step": 1034 }, { "epoch": 0.13166263834117797, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.841480731964111, "learning_rate": 4.383213225943196e-07, "loss": 0.5029, "mean_token_accuracy": 0.8365890383720398, "num_tokens": 39441303.0, "step": 1035 }, { "epoch": 0.13178984861976847, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.828182697296143, "learning_rate": 4.3874523103009746e-07, "loss": 0.4477, "mean_token_accuracy": 0.855778694152832, "num_tokens": 39477359.0, "step": 1036 }, { "epoch": 0.13191705889835897, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 5.830970764160156, "learning_rate": 4.3916913946587536e-07, "loss": 0.4576, "mean_token_accuracy": 0.8483437895774841, "num_tokens": 39511687.0, "step": 1037 }, { "epoch": 0.1320442691769495, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.825500011444092, "learning_rate": 4.3959304790165325e-07, "loss": 0.4726, "mean_token_accuracy": 0.8446031212806702, "num_tokens": 39547453.0, "step": 1038 }, { "epoch": 0.13217147945554, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 5.810396671295166, "learning_rate": 4.400169563374311e-07, "loss": 0.4898, "mean_token_accuracy": 0.8412513136863708, "num_tokens": 39588934.0, "step": 1039 }, { "epoch": 0.13229868973413053, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.871128082275391, "learning_rate": 4.4044086477320895e-07, "loss": 0.5807, "mean_token_accuracy": 0.8188813328742981, "num_tokens": 39627638.0, "step": 1040 }, { "epoch": 0.13242590001272103, "ewc_loss": 0.004638671875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 5.868357181549072, "learning_rate": 4.4086477320898685e-07, "loss": 0.4885, "mean_token_accuracy": 0.8403201103210449, "num_tokens": 39661696.0, "step": 1041 }, { "epoch": 0.13255311029131153, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.819796562194824, "learning_rate": 4.4128868164476474e-07, "loss": 0.4979, "mean_token_accuracy": 0.8440483808517456, "num_tokens": 39702357.0, "step": 1042 }, { "epoch": 0.13268032056990206, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.8268513679504395, "learning_rate": 4.417125900805426e-07, "loss": 0.4816, "mean_token_accuracy": 0.845038890838623, "num_tokens": 39744960.0, "step": 1043 }, { "epoch": 0.13280753084849256, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.7848949432373047e-06, "grad_norm": 5.86515474319458, "learning_rate": 4.4213649851632044e-07, "loss": 0.5398, "mean_token_accuracy": 0.8306068181991577, "num_tokens": 39785394.0, "step": 1044 }, { "epoch": 0.13293474112708306, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 5.867171287536621, "learning_rate": 4.4256040695209834e-07, "loss": 0.521, "mean_token_accuracy": 0.8355243802070618, "num_tokens": 39822698.0, "step": 1045 }, { "epoch": 0.1330619514056736, "ewc_loss": 0.004638671875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 5.905270576477051, "learning_rate": 4.429843153878762e-07, "loss": 0.5175, "mean_token_accuracy": 0.8329452276229858, "num_tokens": 39856874.0, "step": 1046 }, { "epoch": 0.1331891616842641, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.869548320770264, "learning_rate": 4.434082238236541e-07, "loss": 0.5492, "mean_token_accuracy": 0.8232961893081665, "num_tokens": 39899683.0, "step": 1047 }, { "epoch": 0.1333163719628546, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.830341815948486, "learning_rate": 4.4383213225943193e-07, "loss": 0.5278, "mean_token_accuracy": 0.8326973915100098, "num_tokens": 39940565.0, "step": 1048 }, { "epoch": 0.13344358224144512, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.7997961044311523e-06, "grad_norm": 5.883054733276367, "learning_rate": 4.442560406952098e-07, "loss": 0.5279, "mean_token_accuracy": 0.8291439414024353, "num_tokens": 39979010.0, "step": 1049 }, { "epoch": 0.13357079252003562, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.7848949432373047e-06, "grad_norm": 5.810348987579346, "learning_rate": 4.4467994913098767e-07, "loss": 0.4757, "mean_token_accuracy": 0.8467705249786377, "num_tokens": 40021151.0, "step": 1050 }, { "epoch": 0.13369800279862612, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.7997961044311523e-06, "grad_norm": 5.931797981262207, "learning_rate": 4.4510385756676557e-07, "loss": 0.5362, "mean_token_accuracy": 0.8266159296035767, "num_tokens": 40059324.0, "step": 1051 }, { "epoch": 0.13382521307721665, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.8131279945373535, "learning_rate": 4.455277660025434e-07, "loss": 0.4917, "mean_token_accuracy": 0.8396729826927185, "num_tokens": 40104577.0, "step": 1052 }, { "epoch": 0.13395242335580715, "ewc_loss": 0.004638671875, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 5.881365776062012, "learning_rate": 4.459516744383213e-07, "loss": 0.5151, "mean_token_accuracy": 0.8309890627861023, "num_tokens": 40139734.0, "step": 1053 }, { "epoch": 0.13407963363439765, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.907334327697754, "learning_rate": 4.4637558287409916e-07, "loss": 0.5133, "mean_token_accuracy": 0.8344835638999939, "num_tokens": 40180724.0, "step": 1054 }, { "epoch": 0.13420684391298818, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.843075275421143, "learning_rate": 4.4679949130987706e-07, "loss": 0.5127, "mean_token_accuracy": 0.8404582738876343, "num_tokens": 40220475.0, "step": 1055 }, { "epoch": 0.13433405419157868, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.861129283905029, "learning_rate": 4.472233997456549e-07, "loss": 0.5212, "mean_token_accuracy": 0.8326402902603149, "num_tokens": 40258208.0, "step": 1056 }, { "epoch": 0.13446126447016918, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.872050762176514, "learning_rate": 4.476473081814328e-07, "loss": 0.5068, "mean_token_accuracy": 0.8353776335716248, "num_tokens": 40299127.0, "step": 1057 }, { "epoch": 0.1345884747487597, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.846963882446289, "learning_rate": 4.4807121661721065e-07, "loss": 0.4659, "mean_token_accuracy": 0.8472751975059509, "num_tokens": 40339331.0, "step": 1058 }, { "epoch": 0.1347156850273502, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.889590740203857, "learning_rate": 4.4849512505298855e-07, "loss": 0.5089, "mean_token_accuracy": 0.835767924785614, "num_tokens": 40379386.0, "step": 1059 }, { "epoch": 0.1348428953059407, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.876661777496338, "learning_rate": 4.489190334887664e-07, "loss": 0.5458, "mean_token_accuracy": 0.8305041790008545, "num_tokens": 40419868.0, "step": 1060 }, { "epoch": 0.13497010558453124, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.7848949432373047e-06, "grad_norm": 5.914413928985596, "learning_rate": 4.493429419245443e-07, "loss": 0.5146, "mean_token_accuracy": 0.8366990685462952, "num_tokens": 40455346.0, "step": 1061 }, { "epoch": 0.13509731586312174, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.892031192779541, "learning_rate": 4.4976685036032214e-07, "loss": 0.473, "mean_token_accuracy": 0.8449918031692505, "num_tokens": 40490087.0, "step": 1062 }, { "epoch": 0.13522452614171224, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.9109625816345215, "learning_rate": 4.5019075879610004e-07, "loss": 0.5309, "mean_token_accuracy": 0.8266227841377258, "num_tokens": 40524799.0, "step": 1063 }, { "epoch": 0.13535173642030277, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.912869930267334, "learning_rate": 4.506146672318779e-07, "loss": 0.5038, "mean_token_accuracy": 0.8370367288589478, "num_tokens": 40563740.0, "step": 1064 }, { "epoch": 0.13547894669889327, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.904908180236816, "learning_rate": 4.5103857566765573e-07, "loss": 0.5577, "mean_token_accuracy": 0.826768696308136, "num_tokens": 40604743.0, "step": 1065 }, { "epoch": 0.1356061569774838, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.91807746887207, "learning_rate": 4.5146248410343363e-07, "loss": 0.4628, "mean_token_accuracy": 0.8498892188072205, "num_tokens": 40640909.0, "step": 1066 }, { "epoch": 0.1357333672560743, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.881168842315674, "learning_rate": 4.5188639253921153e-07, "loss": 0.4715, "mean_token_accuracy": 0.8473424315452576, "num_tokens": 40678930.0, "step": 1067 }, { "epoch": 0.1358605775346648, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.9364447593688965, "learning_rate": 4.523103009749894e-07, "loss": 0.4658, "mean_token_accuracy": 0.846892774105072, "num_tokens": 40715609.0, "step": 1068 }, { "epoch": 0.13598778781325532, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.879117012023926, "learning_rate": 4.527342094107672e-07, "loss": 0.4903, "mean_token_accuracy": 0.8408386707305908, "num_tokens": 40760533.0, "step": 1069 }, { "epoch": 0.13611499809184582, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.933991432189941, "learning_rate": 4.531581178465451e-07, "loss": 0.5173, "mean_token_accuracy": 0.8343040943145752, "num_tokens": 40800039.0, "step": 1070 }, { "epoch": 0.13624220837043632, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.943483352661133, "learning_rate": 4.53582026282323e-07, "loss": 0.4805, "mean_token_accuracy": 0.8450367450714111, "num_tokens": 40840207.0, "step": 1071 }, { "epoch": 0.13636941864902685, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.950591564178467, "learning_rate": 4.5400593471810087e-07, "loss": 0.5624, "mean_token_accuracy": 0.819817066192627, "num_tokens": 40879559.0, "step": 1072 }, { "epoch": 0.13649662892761735, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.928666591644287, "learning_rate": 4.544298431538787e-07, "loss": 0.4897, "mean_token_accuracy": 0.8423287868499756, "num_tokens": 40919608.0, "step": 1073 }, { "epoch": 0.13662383920620785, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.958032608032227, "learning_rate": 4.548537515896566e-07, "loss": 0.5498, "mean_token_accuracy": 0.8238303661346436, "num_tokens": 40953648.0, "step": 1074 }, { "epoch": 0.13675104948479838, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.970132350921631, "learning_rate": 4.552776600254345e-07, "loss": 0.5221, "mean_token_accuracy": 0.8334953784942627, "num_tokens": 40989675.0, "step": 1075 }, { "epoch": 0.13687825976338888, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.947009086608887, "learning_rate": 4.5570156846121236e-07, "loss": 0.4735, "mean_token_accuracy": 0.8475117683410645, "num_tokens": 41027992.0, "step": 1076 }, { "epoch": 0.13700547004197938, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.940152168273926, "learning_rate": 4.561254768969902e-07, "loss": 0.4725, "mean_token_accuracy": 0.8441969156265259, "num_tokens": 41066925.0, "step": 1077 }, { "epoch": 0.1371326803205699, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.898688316345215, "learning_rate": 4.565493853327681e-07, "loss": 0.4817, "mean_token_accuracy": 0.8444653749465942, "num_tokens": 41105090.0, "step": 1078 }, { "epoch": 0.1372598905991604, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.9707512855529785, "learning_rate": 4.56973293768546e-07, "loss": 0.5242, "mean_token_accuracy": 0.8354917764663696, "num_tokens": 41141924.0, "step": 1079 }, { "epoch": 0.1373871008777509, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 6.011508464813232, "learning_rate": 4.573972022043238e-07, "loss": 0.5853, "mean_token_accuracy": 0.8135610818862915, "num_tokens": 41181512.0, "step": 1080 }, { "epoch": 0.13751431115634144, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.8797712326049805, "learning_rate": 4.578211106401017e-07, "loss": 0.5335, "mean_token_accuracy": 0.8285824060440063, "num_tokens": 41224015.0, "step": 1081 }, { "epoch": 0.13764152143493194, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.951341152191162, "learning_rate": 4.582450190758796e-07, "loss": 0.5352, "mean_token_accuracy": 0.8307178616523743, "num_tokens": 41265568.0, "step": 1082 }, { "epoch": 0.13776873171352244, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.9285688400268555, "learning_rate": 4.586689275116575e-07, "loss": 0.4492, "mean_token_accuracy": 0.8544108271598816, "num_tokens": 41303595.0, "step": 1083 }, { "epoch": 0.13789594199211297, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.921864986419678, "learning_rate": 4.590928359474353e-07, "loss": 0.5525, "mean_token_accuracy": 0.8248560428619385, "num_tokens": 41340119.0, "step": 1084 }, { "epoch": 0.13802315227070347, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.994368553161621, "learning_rate": 4.595167443832132e-07, "loss": 0.5038, "mean_token_accuracy": 0.8408005237579346, "num_tokens": 41380184.0, "step": 1085 }, { "epoch": 0.13815036254929397, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.896311283111572, "learning_rate": 4.599406528189911e-07, "loss": 0.4472, "mean_token_accuracy": 0.8539001941680908, "num_tokens": 41419038.0, "step": 1086 }, { "epoch": 0.1382775728278845, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.95438289642334, "learning_rate": 4.60364561254769e-07, "loss": 0.4902, "mean_token_accuracy": 0.8387380242347717, "num_tokens": 41455767.0, "step": 1087 }, { "epoch": 0.138404783106475, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.989275932312012, "learning_rate": 4.607884696905468e-07, "loss": 0.4875, "mean_token_accuracy": 0.8409921526908875, "num_tokens": 41497493.0, "step": 1088 }, { "epoch": 0.1385319933850655, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.951836109161377, "learning_rate": 4.612123781263247e-07, "loss": 0.519, "mean_token_accuracy": 0.8327101469039917, "num_tokens": 41542940.0, "step": 1089 }, { "epoch": 0.13865920366365603, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.935186862945557, "learning_rate": 4.616362865621026e-07, "loss": 0.4192, "mean_token_accuracy": 0.8604480028152466, "num_tokens": 41579302.0, "step": 1090 }, { "epoch": 0.13878641394224653, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 6.022659778594971, "learning_rate": 4.620601949978805e-07, "loss": 0.5177, "mean_token_accuracy": 0.8347984552383423, "num_tokens": 41617555.0, "step": 1091 }, { "epoch": 0.13891362422083706, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 6.040883541107178, "learning_rate": 4.6248410343365827e-07, "loss": 0.5692, "mean_token_accuracy": 0.8181538581848145, "num_tokens": 41654660.0, "step": 1092 }, { "epoch": 0.13904083449942756, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.9918622970581055, "learning_rate": 4.6290801186943617e-07, "loss": 0.4763, "mean_token_accuracy": 0.8447736501693726, "num_tokens": 41693773.0, "step": 1093 }, { "epoch": 0.13916804477801806, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.9552459716796875, "learning_rate": 4.6333192030521407e-07, "loss": 0.4903, "mean_token_accuracy": 0.8437521457672119, "num_tokens": 41728864.0, "step": 1094 }, { "epoch": 0.13929525505660859, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.920643329620361, "learning_rate": 4.6375582874099196e-07, "loss": 0.4648, "mean_token_accuracy": 0.849625825881958, "num_tokens": 41770596.0, "step": 1095 }, { "epoch": 0.1394224653351991, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.967784881591797, "learning_rate": 4.6417973717676976e-07, "loss": 0.4827, "mean_token_accuracy": 0.8465942144393921, "num_tokens": 41806585.0, "step": 1096 }, { "epoch": 0.1395496756137896, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.9516096115112305, "learning_rate": 4.6460364561254766e-07, "loss": 0.5112, "mean_token_accuracy": 0.834456741809845, "num_tokens": 41845488.0, "step": 1097 }, { "epoch": 0.13967688589238011, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 6.007695198059082, "learning_rate": 4.6502755404832556e-07, "loss": 0.4779, "mean_token_accuracy": 0.8440968990325928, "num_tokens": 41880674.0, "step": 1098 }, { "epoch": 0.13980409617097062, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.039077281951904, "learning_rate": 4.654514624841034e-07, "loss": 0.4923, "mean_token_accuracy": 0.8401867151260376, "num_tokens": 41913020.0, "step": 1099 }, { "epoch": 0.13993130644956112, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 6.032159328460693, "learning_rate": 4.6587537091988125e-07, "loss": 0.4976, "mean_token_accuracy": 0.8402493000030518, "num_tokens": 41948483.0, "step": 1100 }, { "epoch": 0.14005851672815164, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 6.008721828460693, "learning_rate": 4.6629927935565915e-07, "loss": 0.4698, "mean_token_accuracy": 0.8474698066711426, "num_tokens": 41985659.0, "step": 1101 }, { "epoch": 0.14018572700674214, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.986001014709473, "learning_rate": 4.6672318779143705e-07, "loss": 0.5652, "mean_token_accuracy": 0.8198689222335815, "num_tokens": 42027746.0, "step": 1102 }, { "epoch": 0.14031293728533265, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 6.05239200592041, "learning_rate": 4.671470962272149e-07, "loss": 0.521, "mean_token_accuracy": 0.8318740725517273, "num_tokens": 42060903.0, "step": 1103 }, { "epoch": 0.14044014756392317, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.981773376464844, "learning_rate": 4.6757100466299274e-07, "loss": 0.5175, "mean_token_accuracy": 0.8370128273963928, "num_tokens": 42102722.0, "step": 1104 }, { "epoch": 0.14056735784251367, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.94321870803833, "learning_rate": 4.6799491309877064e-07, "loss": 0.5261, "mean_token_accuracy": 0.828667402267456, "num_tokens": 42143418.0, "step": 1105 }, { "epoch": 0.14069456812110417, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.042679309844971, "learning_rate": 4.6841882153454854e-07, "loss": 0.4652, "mean_token_accuracy": 0.8458895683288574, "num_tokens": 42177313.0, "step": 1106 }, { "epoch": 0.1408217783996947, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.0168681144714355, "learning_rate": 4.688427299703264e-07, "loss": 0.4456, "mean_token_accuracy": 0.8522093296051025, "num_tokens": 42214158.0, "step": 1107 }, { "epoch": 0.1409489886782852, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.976102828979492, "learning_rate": 4.6926663840610423e-07, "loss": 0.5042, "mean_token_accuracy": 0.8391220569610596, "num_tokens": 42254189.0, "step": 1108 }, { "epoch": 0.1410761989568757, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.0310750007629395, "learning_rate": 4.6969054684188213e-07, "loss": 0.5254, "mean_token_accuracy": 0.8280065655708313, "num_tokens": 42296853.0, "step": 1109 }, { "epoch": 0.14120340923546623, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.040484428405762, "learning_rate": 4.7011445527766003e-07, "loss": 0.4969, "mean_token_accuracy": 0.8367669582366943, "num_tokens": 42331983.0, "step": 1110 }, { "epoch": 0.14133061951405673, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.715557098388672e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.01849365234375, "learning_rate": 4.7053836371343787e-07, "loss": 0.5059, "mean_token_accuracy": 0.8387454152107239, "num_tokens": 42370496.0, "step": 1111 }, { "epoch": 0.14145782979264723, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.019172191619873, "learning_rate": 4.709622721492157e-07, "loss": 0.4838, "mean_token_accuracy": 0.8442601561546326, "num_tokens": 42406489.0, "step": 1112 }, { "epoch": 0.14158504007123776, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.016329765319824, "learning_rate": 4.713861805849936e-07, "loss": 0.4839, "mean_token_accuracy": 0.8426024317741394, "num_tokens": 42440679.0, "step": 1113 }, { "epoch": 0.14171225034982826, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.066356658935547, "learning_rate": 4.718100890207715e-07, "loss": 0.5565, "mean_token_accuracy": 0.8212366104125977, "num_tokens": 42476699.0, "step": 1114 }, { "epoch": 0.1418394606284188, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.044421195983887, "learning_rate": 4.7223399745654936e-07, "loss": 0.5123, "mean_token_accuracy": 0.8350310921669006, "num_tokens": 42513348.0, "step": 1115 }, { "epoch": 0.1419666709070093, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.029896259307861, "learning_rate": 4.726579058923272e-07, "loss": 0.505, "mean_token_accuracy": 0.8398175835609436, "num_tokens": 42550633.0, "step": 1116 }, { "epoch": 0.1420938811855998, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.0058746337890625, "learning_rate": 4.730818143281051e-07, "loss": 0.5292, "mean_token_accuracy": 0.8324183225631714, "num_tokens": 42593649.0, "step": 1117 }, { "epoch": 0.14222109146419032, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.00727653503418, "learning_rate": 4.7350572276388295e-07, "loss": 0.4759, "mean_token_accuracy": 0.8456957340240479, "num_tokens": 42632379.0, "step": 1118 }, { "epoch": 0.14234830174278082, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 6.0737104415893555, "learning_rate": 4.7392963119966085e-07, "loss": 0.5024, "mean_token_accuracy": 0.8358845114707947, "num_tokens": 42670319.0, "step": 1119 }, { "epoch": 0.14247551202137132, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 6.020637035369873, "learning_rate": 4.7435353963543875e-07, "loss": 0.4955, "mean_token_accuracy": 0.8426500558853149, "num_tokens": 42712680.0, "step": 1120 }, { "epoch": 0.14260272229996185, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 6.059811592102051, "learning_rate": 4.747774480712166e-07, "loss": 0.5127, "mean_token_accuracy": 0.8374239206314087, "num_tokens": 42748241.0, "step": 1121 }, { "epoch": 0.14272993257855235, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.0736985206604, "learning_rate": 4.7520135650699444e-07, "loss": 0.5329, "mean_token_accuracy": 0.8297255039215088, "num_tokens": 42783208.0, "step": 1122 }, { "epoch": 0.14285714285714285, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 6.10005521774292, "learning_rate": 4.7562526494277234e-07, "loss": 0.5079, "mean_token_accuracy": 0.8384791612625122, "num_tokens": 42818570.0, "step": 1123 }, { "epoch": 0.14298435313573338, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.045956134796143, "learning_rate": 4.7604917337855024e-07, "loss": 0.4461, "mean_token_accuracy": 0.8524978756904602, "num_tokens": 42852891.0, "step": 1124 }, { "epoch": 0.14311156341432388, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.066133499145508, "learning_rate": 4.764730818143281e-07, "loss": 0.4961, "mean_token_accuracy": 0.8331838846206665, "num_tokens": 42887359.0, "step": 1125 }, { "epoch": 0.14323877369291438, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.077966690063477, "learning_rate": 4.768969902501059e-07, "loss": 0.5488, "mean_token_accuracy": 0.8240115642547607, "num_tokens": 42929308.0, "step": 1126 }, { "epoch": 0.1433659839715049, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.023763179779053, "learning_rate": 4.773208986858838e-07, "loss": 0.4987, "mean_token_accuracy": 0.8364932537078857, "num_tokens": 42969373.0, "step": 1127 }, { "epoch": 0.1434931942500954, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.059108257293701, "learning_rate": 4.777448071216617e-07, "loss": 0.5039, "mean_token_accuracy": 0.8378851413726807, "num_tokens": 43010133.0, "step": 1128 }, { "epoch": 0.1436204045286859, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.072702407836914, "learning_rate": 4.781687155574396e-07, "loss": 0.5286, "mean_token_accuracy": 0.8293555974960327, "num_tokens": 43053743.0, "step": 1129 }, { "epoch": 0.14374761480727644, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.033459186553955, "learning_rate": 4.785926239932175e-07, "loss": 0.4696, "mean_token_accuracy": 0.84751296043396, "num_tokens": 43094125.0, "step": 1130 }, { "epoch": 0.14387482508586694, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.0824875831604, "learning_rate": 4.790165324289953e-07, "loss": 0.4959, "mean_token_accuracy": 0.8375097513198853, "num_tokens": 43132399.0, "step": 1131 }, { "epoch": 0.14400203536445744, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.156536102294922, "learning_rate": 4.794404408647732e-07, "loss": 0.5104, "mean_token_accuracy": 0.8372979760169983, "num_tokens": 43169036.0, "step": 1132 }, { "epoch": 0.14412924564304797, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.048627853393555, "learning_rate": 4.798643493005511e-07, "loss": 0.4413, "mean_token_accuracy": 0.8545522689819336, "num_tokens": 43209397.0, "step": 1133 }, { "epoch": 0.14425645592163847, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 6.071885108947754, "learning_rate": 4.80288257736329e-07, "loss": 0.5385, "mean_token_accuracy": 0.8287335634231567, "num_tokens": 43246788.0, "step": 1134 }, { "epoch": 0.14438366620022897, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.110028266906738, "learning_rate": 4.807121661721068e-07, "loss": 0.4503, "mean_token_accuracy": 0.853098452091217, "num_tokens": 43282774.0, "step": 1135 }, { "epoch": 0.1445108764788195, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.081356048583984, "learning_rate": 4.811360746078847e-07, "loss": 0.5065, "mean_token_accuracy": 0.8402565121650696, "num_tokens": 43322303.0, "step": 1136 }, { "epoch": 0.14463808675741, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.139896392822266, "learning_rate": 4.815599830436625e-07, "loss": 0.5002, "mean_token_accuracy": 0.8378446698188782, "num_tokens": 43354611.0, "step": 1137 }, { "epoch": 0.1447652970360005, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.107308387756348, "learning_rate": 4.819838914794405e-07, "loss": 0.4819, "mean_token_accuracy": 0.8444901704788208, "num_tokens": 43390640.0, "step": 1138 }, { "epoch": 0.14489250731459102, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.118914604187012, "learning_rate": 4.824077999152183e-07, "loss": 0.5017, "mean_token_accuracy": 0.8340405821800232, "num_tokens": 43424513.0, "step": 1139 }, { "epoch": 0.14501971759318152, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.094738483428955, "learning_rate": 4.828317083509962e-07, "loss": 0.5078, "mean_token_accuracy": 0.8350386023521423, "num_tokens": 43461699.0, "step": 1140 }, { "epoch": 0.14514692787177205, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.177622318267822, "learning_rate": 4.83255616786774e-07, "loss": 0.4815, "mean_token_accuracy": 0.8485947847366333, "num_tokens": 43494681.0, "step": 1141 }, { "epoch": 0.14527413815036255, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.078976154327393, "learning_rate": 4.83679525222552e-07, "loss": 0.4577, "mean_token_accuracy": 0.8514358401298523, "num_tokens": 43531318.0, "step": 1142 }, { "epoch": 0.14540134842895305, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.105060577392578, "learning_rate": 4.841034336583298e-07, "loss": 0.4738, "mean_token_accuracy": 0.850806713104248, "num_tokens": 43569991.0, "step": 1143 }, { "epoch": 0.14552855870754358, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.099667072296143, "learning_rate": 4.845273420941076e-07, "loss": 0.462, "mean_token_accuracy": 0.8518836498260498, "num_tokens": 43609978.0, "step": 1144 }, { "epoch": 0.14565576898613408, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.155895709991455, "learning_rate": 4.849512505298855e-07, "loss": 0.5097, "mean_token_accuracy": 0.8303203582763672, "num_tokens": 43640419.0, "step": 1145 }, { "epoch": 0.14578297926472458, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.1103081703186035, "learning_rate": 4.853751589656634e-07, "loss": 0.4715, "mean_token_accuracy": 0.847861647605896, "num_tokens": 43682408.0, "step": 1146 }, { "epoch": 0.1459101895433151, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.086642742156982, "learning_rate": 4.857990674014413e-07, "loss": 0.5093, "mean_token_accuracy": 0.834699273109436, "num_tokens": 43726158.0, "step": 1147 }, { "epoch": 0.1460373998219056, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.130925178527832, "learning_rate": 4.862229758372191e-07, "loss": 0.539, "mean_token_accuracy": 0.8238727450370789, "num_tokens": 43766466.0, "step": 1148 }, { "epoch": 0.1461646101004961, "ewc_loss": 0.005126953125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 6.147735595703125, "learning_rate": 4.86646884272997e-07, "loss": 0.5264, "mean_token_accuracy": 0.8285349011421204, "num_tokens": 43807255.0, "step": 1149 }, { "epoch": 0.14629182037908664, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.135196685791016, "learning_rate": 4.870707927087749e-07, "loss": 0.5048, "mean_token_accuracy": 0.8371759057044983, "num_tokens": 43845419.0, "step": 1150 }, { "epoch": 0.14641903065767714, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.150813102722168, "learning_rate": 4.874947011445528e-07, "loss": 0.5132, "mean_token_accuracy": 0.8364908695220947, "num_tokens": 43882942.0, "step": 1151 }, { "epoch": 0.14654624093626764, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.091933250427246, "learning_rate": 4.879186095803306e-07, "loss": 0.4338, "mean_token_accuracy": 0.8584513068199158, "num_tokens": 43923109.0, "step": 1152 }, { "epoch": 0.14667345121485817, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.140272617340088, "learning_rate": 4.883425180161085e-07, "loss": 0.4711, "mean_token_accuracy": 0.8459732532501221, "num_tokens": 43959752.0, "step": 1153 }, { "epoch": 0.14680066149344867, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.157310485839844, "learning_rate": 4.887664264518864e-07, "loss": 0.5182, "mean_token_accuracy": 0.8376511931419373, "num_tokens": 43996247.0, "step": 1154 }, { "epoch": 0.14692787177203917, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.083136558532715, "learning_rate": 4.891903348876643e-07, "loss": 0.5014, "mean_token_accuracy": 0.8410637974739075, "num_tokens": 44035751.0, "step": 1155 }, { "epoch": 0.1470550820506297, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.10646915435791, "learning_rate": 4.896142433234421e-07, "loss": 0.5358, "mean_token_accuracy": 0.8336150646209717, "num_tokens": 44073225.0, "step": 1156 }, { "epoch": 0.1471822923292202, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.130025386810303, "learning_rate": 4.9003815175922e-07, "loss": 0.5387, "mean_token_accuracy": 0.8327957391738892, "num_tokens": 44111761.0, "step": 1157 }, { "epoch": 0.1473095026078107, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.100727081298828, "learning_rate": 4.904620601949979e-07, "loss": 0.4895, "mean_token_accuracy": 0.841663122177124, "num_tokens": 44151311.0, "step": 1158 }, { "epoch": 0.14743671288640123, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.194584846496582, "learning_rate": 4.908859686307758e-07, "loss": 0.5343, "mean_token_accuracy": 0.83089679479599, "num_tokens": 44184623.0, "step": 1159 }, { "epoch": 0.14756392316499173, "ewc_loss": 0.005126953125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 6.1354475021362305, "learning_rate": 4.913098770665536e-07, "loss": 0.5202, "mean_token_accuracy": 0.8362781405448914, "num_tokens": 44221753.0, "step": 1160 }, { "epoch": 0.14769113344358223, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.14157247543335, "learning_rate": 4.917337855023314e-07, "loss": 0.4735, "mean_token_accuracy": 0.8431822657585144, "num_tokens": 44262542.0, "step": 1161 }, { "epoch": 0.14781834372217276, "ewc_loss": 0.005126953125, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.1695451736450195, "learning_rate": 4.921576939381094e-07, "loss": 0.4961, "mean_token_accuracy": 0.8447555303573608, "num_tokens": 44300294.0, "step": 1162 }, { "epoch": 0.14794555400076326, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.14352560043335, "learning_rate": 4.925816023738872e-07, "loss": 0.541, "mean_token_accuracy": 0.8255488872528076, "num_tokens": 44338476.0, "step": 1163 }, { "epoch": 0.14807276427935376, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.2077202796936035, "learning_rate": 4.930055108096651e-07, "loss": 0.4427, "mean_token_accuracy": 0.8563581705093384, "num_tokens": 44367437.0, "step": 1164 }, { "epoch": 0.1481999745579443, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.195346832275391, "learning_rate": 4.934294192454429e-07, "loss": 0.5544, "mean_token_accuracy": 0.822826623916626, "num_tokens": 44402636.0, "step": 1165 }, { "epoch": 0.1483271848365348, "ewc_loss": 0.005126953125, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.164788246154785, "learning_rate": 4.938533276812209e-07, "loss": 0.4967, "mean_token_accuracy": 0.8413326740264893, "num_tokens": 44438187.0, "step": 1166 }, { "epoch": 0.14845439511512531, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 6.102897644042969, "learning_rate": 4.942772361169987e-07, "loss": 0.4194, "mean_token_accuracy": 0.8585610389709473, "num_tokens": 44478345.0, "step": 1167 }, { "epoch": 0.14858160539371582, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.125441551208496, "learning_rate": 4.947011445527766e-07, "loss": 0.4989, "mean_token_accuracy": 0.8391841053962708, "num_tokens": 44520272.0, "step": 1168 }, { "epoch": 0.14870881567230632, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.136446475982666, "learning_rate": 4.951250529885544e-07, "loss": 0.4508, "mean_token_accuracy": 0.8557511568069458, "num_tokens": 44561058.0, "step": 1169 }, { "epoch": 0.14883602595089684, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.212655067443848, "learning_rate": 4.955489614243324e-07, "loss": 0.5837, "mean_token_accuracy": 0.811421275138855, "num_tokens": 44599371.0, "step": 1170 }, { "epoch": 0.14896323622948734, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.1766438484191895, "learning_rate": 4.959728698601102e-07, "loss": 0.4947, "mean_token_accuracy": 0.8318784236907959, "num_tokens": 44640499.0, "step": 1171 }, { "epoch": 0.14909044650807785, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 6.108429431915283, "learning_rate": 4.963967782958881e-07, "loss": 0.4814, "mean_token_accuracy": 0.8449016809463501, "num_tokens": 44683164.0, "step": 1172 }, { "epoch": 0.14921765678666837, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.193994522094727, "learning_rate": 4.968206867316659e-07, "loss": 0.4932, "mean_token_accuracy": 0.8388310074806213, "num_tokens": 44721938.0, "step": 1173 }, { "epoch": 0.14934486706525887, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.2396321296691895, "learning_rate": 4.972445951674439e-07, "loss": 0.4643, "mean_token_accuracy": 0.847274661064148, "num_tokens": 44753506.0, "step": 1174 }, { "epoch": 0.14947207734384937, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.198103427886963, "learning_rate": 4.976685036032216e-07, "loss": 0.5556, "mean_token_accuracy": 0.8223271369934082, "num_tokens": 44792170.0, "step": 1175 }, { "epoch": 0.1495992876224399, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.147575855255127, "learning_rate": 4.980924120389996e-07, "loss": 0.4324, "mean_token_accuracy": 0.8596873879432678, "num_tokens": 44830733.0, "step": 1176 }, { "epoch": 0.1497264979010304, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.173964500427246, "learning_rate": 4.985163204747774e-07, "loss": 0.4457, "mean_token_accuracy": 0.8551669716835022, "num_tokens": 44869743.0, "step": 1177 }, { "epoch": 0.1498537081796209, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.193195343017578, "learning_rate": 4.989402289105554e-07, "loss": 0.4925, "mean_token_accuracy": 0.8457858562469482, "num_tokens": 44913358.0, "step": 1178 }, { "epoch": 0.14998091845821143, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.240006923675537, "learning_rate": 4.993641373463331e-07, "loss": 0.5521, "mean_token_accuracy": 0.8229309916496277, "num_tokens": 44949963.0, "step": 1179 }, { "epoch": 0.15010812873680193, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.204779148101807, "learning_rate": 4.997880457821111e-07, "loss": 0.5066, "mean_token_accuracy": 0.8350269794464111, "num_tokens": 44988569.0, "step": 1180 }, { "epoch": 0.15023533901539243, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.146470546722412, "learning_rate": 5.002119542178889e-07, "loss": 0.4838, "mean_token_accuracy": 0.8423288464546204, "num_tokens": 45032634.0, "step": 1181 }, { "epoch": 0.15036254929398296, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.218085765838623, "learning_rate": 5.006358626536667e-07, "loss": 0.4523, "mean_token_accuracy": 0.8540387749671936, "num_tokens": 45069685.0, "step": 1182 }, { "epoch": 0.15048975957257346, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.206881999969482, "learning_rate": 5.010597710894446e-07, "loss": 0.4751, "mean_token_accuracy": 0.8457255363464355, "num_tokens": 45107481.0, "step": 1183 }, { "epoch": 0.15061696985116396, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.185901641845703, "learning_rate": 5.014836795252225e-07, "loss": 0.4599, "mean_token_accuracy": 0.85001540184021, "num_tokens": 45145747.0, "step": 1184 }, { "epoch": 0.1507441801297545, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.185728549957275, "learning_rate": 5.019075879610004e-07, "loss": 0.4665, "mean_token_accuracy": 0.8520565032958984, "num_tokens": 45183140.0, "step": 1185 }, { "epoch": 0.150871390408345, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.2147908210754395, "learning_rate": 5.023314963967783e-07, "loss": 0.5005, "mean_token_accuracy": 0.838073194026947, "num_tokens": 45224911.0, "step": 1186 }, { "epoch": 0.1509986006869355, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.224155426025391, "learning_rate": 5.027554048325562e-07, "loss": 0.5552, "mean_token_accuracy": 0.8226988911628723, "num_tokens": 45257388.0, "step": 1187 }, { "epoch": 0.15112581096552602, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.16762638092041, "learning_rate": 5.03179313268334e-07, "loss": 0.5409, "mean_token_accuracy": 0.827178955078125, "num_tokens": 45297876.0, "step": 1188 }, { "epoch": 0.15125302124411652, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.24159049987793, "learning_rate": 5.036032217041119e-07, "loss": 0.516, "mean_token_accuracy": 0.832024872303009, "num_tokens": 45334040.0, "step": 1189 }, { "epoch": 0.15138023152270705, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.171865940093994, "learning_rate": 5.040271301398897e-07, "loss": 0.4997, "mean_token_accuracy": 0.8362027406692505, "num_tokens": 45375985.0, "step": 1190 }, { "epoch": 0.15150744180129755, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.179471015930176, "learning_rate": 5.044510385756676e-07, "loss": 0.4684, "mean_token_accuracy": 0.8451697826385498, "num_tokens": 45419951.0, "step": 1191 }, { "epoch": 0.15163465207988805, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.263296127319336, "learning_rate": 5.048749470114455e-07, "loss": 0.4815, "mean_token_accuracy": 0.844759464263916, "num_tokens": 45456988.0, "step": 1192 }, { "epoch": 0.15176186235847858, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.305363655090332, "learning_rate": 5.052988554472234e-07, "loss": 0.5001, "mean_token_accuracy": 0.836758017539978, "num_tokens": 45488226.0, "step": 1193 }, { "epoch": 0.15188907263706908, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.217455863952637, "learning_rate": 5.057227638830013e-07, "loss": 0.4983, "mean_token_accuracy": 0.8394894599914551, "num_tokens": 45524661.0, "step": 1194 }, { "epoch": 0.15201628291565958, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.275196075439453, "learning_rate": 5.061466723187792e-07, "loss": 0.5007, "mean_token_accuracy": 0.839592695236206, "num_tokens": 45563491.0, "step": 1195 }, { "epoch": 0.1521434931942501, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 6.296469688415527, "learning_rate": 5.065705807545569e-07, "loss": 0.4698, "mean_token_accuracy": 0.8496214151382446, "num_tokens": 45596771.0, "step": 1196 }, { "epoch": 0.1522707034728406, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.1640238761901855, "learning_rate": 5.069944891903349e-07, "loss": 0.4872, "mean_token_accuracy": 0.8421340584754944, "num_tokens": 45640049.0, "step": 1197 }, { "epoch": 0.1523979137514311, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.248464584350586, "learning_rate": 5.074183976261127e-07, "loss": 0.489, "mean_token_accuracy": 0.8407440185546875, "num_tokens": 45677375.0, "step": 1198 }, { "epoch": 0.15252512403002164, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.2624125480651855, "learning_rate": 5.078423060618906e-07, "loss": 0.4929, "mean_token_accuracy": 0.8421453237533569, "num_tokens": 45719253.0, "step": 1199 }, { "epoch": 0.15265233430861214, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.269164562225342, "learning_rate": 5.082662144976685e-07, "loss": 0.4888, "mean_token_accuracy": 0.8387503623962402, "num_tokens": 45756065.0, "step": 1200 }, { "epoch": 0.15277954458720264, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.224458694458008, "learning_rate": 5.086901229334464e-07, "loss": 0.4782, "mean_token_accuracy": 0.8475862741470337, "num_tokens": 45791707.0, "step": 1201 }, { "epoch": 0.15290675486579317, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 6.209017753601074, "learning_rate": 5.091140313692243e-07, "loss": 0.4927, "mean_token_accuracy": 0.8394243717193604, "num_tokens": 45828464.0, "step": 1202 }, { "epoch": 0.15303396514438367, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.219574928283691, "learning_rate": 5.095379398050022e-07, "loss": 0.4756, "mean_token_accuracy": 0.8446881771087646, "num_tokens": 45867046.0, "step": 1203 }, { "epoch": 0.15316117542297417, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 6.296317100524902, "learning_rate": 5.099618482407799e-07, "loss": 0.5062, "mean_token_accuracy": 0.8364729881286621, "num_tokens": 45900530.0, "step": 1204 }, { "epoch": 0.1532883857015647, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.190175533294678, "learning_rate": 5.103857566765578e-07, "loss": 0.4834, "mean_token_accuracy": 0.8441483378410339, "num_tokens": 45940002.0, "step": 1205 }, { "epoch": 0.1534155959801552, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.239543437957764, "learning_rate": 5.108096651123357e-07, "loss": 0.4465, "mean_token_accuracy": 0.854223370552063, "num_tokens": 45975299.0, "step": 1206 }, { "epoch": 0.1535428062587457, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.253157615661621, "learning_rate": 5.112335735481135e-07, "loss": 0.4852, "mean_token_accuracy": 0.8410269021987915, "num_tokens": 46012693.0, "step": 1207 }, { "epoch": 0.15367001653733622, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.175981521606445, "learning_rate": 5.116574819838915e-07, "loss": 0.3913, "mean_token_accuracy": 0.8721171617507935, "num_tokens": 46051558.0, "step": 1208 }, { "epoch": 0.15379722681592672, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.300548553466797, "learning_rate": 5.120813904196693e-07, "loss": 0.4747, "mean_token_accuracy": 0.8471435904502869, "num_tokens": 46087173.0, "step": 1209 }, { "epoch": 0.15392443709451722, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.252926349639893, "learning_rate": 5.125052988554473e-07, "loss": 0.5301, "mean_token_accuracy": 0.8327137231826782, "num_tokens": 46120482.0, "step": 1210 }, { "epoch": 0.15405164737310775, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.224769115447998, "learning_rate": 5.12929207291225e-07, "loss": 0.477, "mean_token_accuracy": 0.8436387777328491, "num_tokens": 46154478.0, "step": 1211 }, { "epoch": 0.15417885765169825, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.245028018951416, "learning_rate": 5.133531157270029e-07, "loss": 0.4545, "mean_token_accuracy": 0.8522144556045532, "num_tokens": 46195012.0, "step": 1212 }, { "epoch": 0.15430606793028875, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.219833850860596, "learning_rate": 5.137770241627808e-07, "loss": 0.4195, "mean_token_accuracy": 0.8624212145805359, "num_tokens": 46230641.0, "step": 1213 }, { "epoch": 0.15443327820887928, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.229718208312988, "learning_rate": 5.142009325985587e-07, "loss": 0.4592, "mean_token_accuracy": 0.8538376092910767, "num_tokens": 46268993.0, "step": 1214 }, { "epoch": 0.15456048848746978, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.244008541107178, "learning_rate": 5.146248410343365e-07, "loss": 0.4922, "mean_token_accuracy": 0.8413923382759094, "num_tokens": 46309310.0, "step": 1215 }, { "epoch": 0.1546876987660603, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.2226738929748535, "learning_rate": 5.150487494701145e-07, "loss": 0.434, "mean_token_accuracy": 0.8581292033195496, "num_tokens": 46352415.0, "step": 1216 }, { "epoch": 0.1548149090446508, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.282905101776123, "learning_rate": 5.154726579058923e-07, "loss": 0.4893, "mean_token_accuracy": 0.839939296245575, "num_tokens": 46387838.0, "step": 1217 }, { "epoch": 0.1549421193232413, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.2473063468933105, "learning_rate": 5.158965663416703e-07, "loss": 0.4845, "mean_token_accuracy": 0.8433252573013306, "num_tokens": 46432669.0, "step": 1218 }, { "epoch": 0.15506932960183184, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.19190788269043, "learning_rate": 5.16320474777448e-07, "loss": 0.4404, "mean_token_accuracy": 0.855932354927063, "num_tokens": 46474383.0, "step": 1219 }, { "epoch": 0.15519653988042234, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.297237396240234, "learning_rate": 5.167443832132259e-07, "loss": 0.4899, "mean_token_accuracy": 0.8411003351211548, "num_tokens": 46514263.0, "step": 1220 }, { "epoch": 0.15532375015901284, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.2970356941223145, "learning_rate": 5.171682916490038e-07, "loss": 0.458, "mean_token_accuracy": 0.8505592942237854, "num_tokens": 46549835.0, "step": 1221 }, { "epoch": 0.15545096043760337, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.286118984222412, "learning_rate": 5.175922000847816e-07, "loss": 0.4449, "mean_token_accuracy": 0.8551504611968994, "num_tokens": 46584811.0, "step": 1222 }, { "epoch": 0.15557817071619387, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.243515968322754, "learning_rate": 5.180161085205595e-07, "loss": 0.4667, "mean_token_accuracy": 0.8491251468658447, "num_tokens": 46621441.0, "step": 1223 }, { "epoch": 0.15570538099478437, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.265459060668945, "learning_rate": 5.184400169563374e-07, "loss": 0.5196, "mean_token_accuracy": 0.8325088024139404, "num_tokens": 46661572.0, "step": 1224 }, { "epoch": 0.1558325912733749, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.258798122406006, "learning_rate": 5.188639253921153e-07, "loss": 0.4945, "mean_token_accuracy": 0.8389711976051331, "num_tokens": 46702157.0, "step": 1225 }, { "epoch": 0.1559598015519654, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.25163459777832, "learning_rate": 5.192878338278932e-07, "loss": 0.4704, "mean_token_accuracy": 0.84714275598526, "num_tokens": 46739119.0, "step": 1226 }, { "epoch": 0.1560870118305559, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.300620079040527, "learning_rate": 5.19711742263671e-07, "loss": 0.5295, "mean_token_accuracy": 0.8306248188018799, "num_tokens": 46783990.0, "step": 1227 }, { "epoch": 0.15621422210914643, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.320226669311523, "learning_rate": 5.201356506994488e-07, "loss": 0.4676, "mean_token_accuracy": 0.842390775680542, "num_tokens": 46816131.0, "step": 1228 }, { "epoch": 0.15634143238773693, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.2728071212768555, "learning_rate": 5.205595591352268e-07, "loss": 0.5505, "mean_token_accuracy": 0.8241318464279175, "num_tokens": 46856453.0, "step": 1229 }, { "epoch": 0.15646864266632743, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.288565635681152, "learning_rate": 5.209834675710046e-07, "loss": 0.4896, "mean_token_accuracy": 0.8416500091552734, "num_tokens": 46897961.0, "step": 1230 }, { "epoch": 0.15659585294491796, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.337296009063721, "learning_rate": 5.214073760067825e-07, "loss": 0.5637, "mean_token_accuracy": 0.8208472728729248, "num_tokens": 46931961.0, "step": 1231 }, { "epoch": 0.15672306322350846, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.29495906829834, "learning_rate": 5.218312844425604e-07, "loss": 0.4844, "mean_token_accuracy": 0.8456091284751892, "num_tokens": 46967822.0, "step": 1232 }, { "epoch": 0.15685027350209896, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.367445945739746, "learning_rate": 5.222551928783383e-07, "loss": 0.4659, "mean_token_accuracy": 0.8483377695083618, "num_tokens": 47004511.0, "step": 1233 }, { "epoch": 0.1569774837806895, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.2939772605896, "learning_rate": 5.226791013141161e-07, "loss": 0.5187, "mean_token_accuracy": 0.8353427648544312, "num_tokens": 47047244.0, "step": 1234 }, { "epoch": 0.15710469405928, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.285640239715576, "learning_rate": 5.23103009749894e-07, "loss": 0.4758, "mean_token_accuracy": 0.8457661867141724, "num_tokens": 47088206.0, "step": 1235 }, { "epoch": 0.1572319043378705, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.270590782165527, "learning_rate": 5.235269181856718e-07, "loss": 0.4844, "mean_token_accuracy": 0.8426474332809448, "num_tokens": 47132733.0, "step": 1236 }, { "epoch": 0.15735911461646102, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.331144332885742, "learning_rate": 5.239508266214498e-07, "loss": 0.5422, "mean_token_accuracy": 0.8232003450393677, "num_tokens": 47166840.0, "step": 1237 }, { "epoch": 0.15748632489505152, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.333258628845215, "learning_rate": 5.243747350572276e-07, "loss": 0.5016, "mean_token_accuracy": 0.8378521203994751, "num_tokens": 47203444.0, "step": 1238 }, { "epoch": 0.15761353517364202, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1205673217773438e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.280689716339111, "learning_rate": 5.247986434930056e-07, "loss": 0.4956, "mean_token_accuracy": 0.8428846001625061, "num_tokens": 47240380.0, "step": 1239 }, { "epoch": 0.15774074545223254, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.332703590393066, "learning_rate": 5.252225519287834e-07, "loss": 0.5555, "mean_token_accuracy": 0.8222392797470093, "num_tokens": 47277886.0, "step": 1240 }, { "epoch": 0.15786795573082305, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.355157852172852, "learning_rate": 5.256464603645613e-07, "loss": 0.4846, "mean_token_accuracy": 0.8467531800270081, "num_tokens": 47315261.0, "step": 1241 }, { "epoch": 0.15799516600941357, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.258774280548096, "learning_rate": 5.260703688003391e-07, "loss": 0.4849, "mean_token_accuracy": 0.8435719013214111, "num_tokens": 47357489.0, "step": 1242 }, { "epoch": 0.15812237628800407, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.3393988609313965, "learning_rate": 5.26494277236117e-07, "loss": 0.5004, "mean_token_accuracy": 0.8390963077545166, "num_tokens": 47389851.0, "step": 1243 }, { "epoch": 0.15824958656659457, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.34238862991333, "learning_rate": 5.269181856718948e-07, "loss": 0.5265, "mean_token_accuracy": 0.8292664289474487, "num_tokens": 47427491.0, "step": 1244 }, { "epoch": 0.1583767968451851, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.329802513122559, "learning_rate": 5.273420941076727e-07, "loss": 0.4672, "mean_token_accuracy": 0.8497462272644043, "num_tokens": 47462790.0, "step": 1245 }, { "epoch": 0.1585040071237756, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.306418418884277, "learning_rate": 5.277660025434506e-07, "loss": 0.4305, "mean_token_accuracy": 0.8585613965988159, "num_tokens": 47502240.0, "step": 1246 }, { "epoch": 0.1586312174023661, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.257391452789307, "learning_rate": 5.281899109792285e-07, "loss": 0.455, "mean_token_accuracy": 0.8517846465110779, "num_tokens": 47542493.0, "step": 1247 }, { "epoch": 0.15875842768095663, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.314043045043945, "learning_rate": 5.286138194150064e-07, "loss": 0.4915, "mean_token_accuracy": 0.842781662940979, "num_tokens": 47583378.0, "step": 1248 }, { "epoch": 0.15888563795954713, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.347504138946533, "learning_rate": 5.290377278507841e-07, "loss": 0.5165, "mean_token_accuracy": 0.8391878008842468, "num_tokens": 47622671.0, "step": 1249 }, { "epoch": 0.15901284823813763, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.460830211639404, "learning_rate": 5.294616362865621e-07, "loss": 0.4754, "mean_token_accuracy": 0.8494881391525269, "num_tokens": 47663774.0, "step": 1250 }, { "epoch": 0.15914005851672816, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.371359348297119, "learning_rate": 5.298855447223399e-07, "loss": 0.5462, "mean_token_accuracy": 0.8338536620140076, "num_tokens": 47706143.0, "step": 1251 }, { "epoch": 0.15926726879531866, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.343655109405518, "learning_rate": 5.303094531581178e-07, "loss": 0.5077, "mean_token_accuracy": 0.8374252915382385, "num_tokens": 47741197.0, "step": 1252 }, { "epoch": 0.15939447907390916, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.337249279022217, "learning_rate": 5.307333615938957e-07, "loss": 0.4312, "mean_token_accuracy": 0.859252393245697, "num_tokens": 47779306.0, "step": 1253 }, { "epoch": 0.1595216893524997, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.420766353607178, "learning_rate": 5.311572700296736e-07, "loss": 0.5164, "mean_token_accuracy": 0.8334348201751709, "num_tokens": 47811056.0, "step": 1254 }, { "epoch": 0.1596488996310902, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.369998455047607, "learning_rate": 5.315811784654515e-07, "loss": 0.4244, "mean_token_accuracy": 0.8574514985084534, "num_tokens": 47847075.0, "step": 1255 }, { "epoch": 0.1597761099096807, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.321204662322998, "learning_rate": 5.320050869012294e-07, "loss": 0.4819, "mean_token_accuracy": 0.8445890545845032, "num_tokens": 47892639.0, "step": 1256 }, { "epoch": 0.15990332018827122, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.345308303833008, "learning_rate": 5.324289953370071e-07, "loss": 0.5397, "mean_token_accuracy": 0.8290544152259827, "num_tokens": 47933522.0, "step": 1257 }, { "epoch": 0.16003053046686172, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.3377556800842285, "learning_rate": 5.328529037727851e-07, "loss": 0.4756, "mean_token_accuracy": 0.8450348377227783, "num_tokens": 47977307.0, "step": 1258 }, { "epoch": 0.16015774074545222, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.416245937347412, "learning_rate": 5.332768122085629e-07, "loss": 0.4731, "mean_token_accuracy": 0.845485508441925, "num_tokens": 48015129.0, "step": 1259 }, { "epoch": 0.16028495102404275, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.412184238433838, "learning_rate": 5.337007206443408e-07, "loss": 0.5079, "mean_token_accuracy": 0.8330034613609314, "num_tokens": 48051904.0, "step": 1260 }, { "epoch": 0.16041216130263325, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.3536787033081055, "learning_rate": 5.341246290801187e-07, "loss": 0.4693, "mean_token_accuracy": 0.850528359413147, "num_tokens": 48094001.0, "step": 1261 }, { "epoch": 0.16053937158122375, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.345480442047119, "learning_rate": 5.345485375158966e-07, "loss": 0.462, "mean_token_accuracy": 0.847417414188385, "num_tokens": 48134124.0, "step": 1262 }, { "epoch": 0.16066658185981428, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.399445056915283, "learning_rate": 5.349724459516745e-07, "loss": 0.4515, "mean_token_accuracy": 0.8516483306884766, "num_tokens": 48170730.0, "step": 1263 }, { "epoch": 0.16079379213840478, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.387270450592041, "learning_rate": 5.353963543874522e-07, "loss": 0.5229, "mean_token_accuracy": 0.8335405588150024, "num_tokens": 48212640.0, "step": 1264 }, { "epoch": 0.1609210024169953, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.441268444061279, "learning_rate": 5.358202628232301e-07, "loss": 0.5002, "mean_token_accuracy": 0.8390688896179199, "num_tokens": 48243136.0, "step": 1265 }, { "epoch": 0.1610482126955858, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.3856306076049805, "learning_rate": 5.36244171259008e-07, "loss": 0.48, "mean_token_accuracy": 0.8452456593513489, "num_tokens": 48282905.0, "step": 1266 }, { "epoch": 0.1611754229741763, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.424999713897705, "learning_rate": 5.366680796947859e-07, "loss": 0.518, "mean_token_accuracy": 0.8358765244483948, "num_tokens": 48323325.0, "step": 1267 }, { "epoch": 0.16130263325276684, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.4602370262146, "learning_rate": 5.370919881305637e-07, "loss": 0.4954, "mean_token_accuracy": 0.8439378142356873, "num_tokens": 48358400.0, "step": 1268 }, { "epoch": 0.16142984353135734, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.416213512420654, "learning_rate": 5.375158965663417e-07, "loss": 0.4973, "mean_token_accuracy": 0.8416191339492798, "num_tokens": 48393018.0, "step": 1269 }, { "epoch": 0.16155705380994784, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.380592346191406, "learning_rate": 5.379398050021195e-07, "loss": 0.4847, "mean_token_accuracy": 0.844409704208374, "num_tokens": 48432952.0, "step": 1270 }, { "epoch": 0.16168426408853837, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.408500671386719, "learning_rate": 5.383637134378975e-07, "loss": 0.4725, "mean_token_accuracy": 0.844440221786499, "num_tokens": 48466913.0, "step": 1271 }, { "epoch": 0.16181147436712887, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.425894260406494, "learning_rate": 5.387876218736752e-07, "loss": 0.5529, "mean_token_accuracy": 0.8196007013320923, "num_tokens": 48512658.0, "step": 1272 }, { "epoch": 0.16193868464571937, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.368729591369629, "learning_rate": 5.392115303094531e-07, "loss": 0.4499, "mean_token_accuracy": 0.8527109026908875, "num_tokens": 48552283.0, "step": 1273 }, { "epoch": 0.1620658949243099, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.400202751159668, "learning_rate": 5.39635438745231e-07, "loss": 0.4295, "mean_token_accuracy": 0.8597226738929749, "num_tokens": 48590504.0, "step": 1274 }, { "epoch": 0.1621931052029004, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.439131736755371, "learning_rate": 5.400593471810089e-07, "loss": 0.518, "mean_token_accuracy": 0.8332165479660034, "num_tokens": 48629993.0, "step": 1275 }, { "epoch": 0.1623203154814909, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.4572319984436035, "learning_rate": 5.404832556167867e-07, "loss": 0.5209, "mean_token_accuracy": 0.8335386514663696, "num_tokens": 48665074.0, "step": 1276 }, { "epoch": 0.16244752576008142, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.403236389160156, "learning_rate": 5.409071640525647e-07, "loss": 0.4788, "mean_token_accuracy": 0.848875880241394, "num_tokens": 48705524.0, "step": 1277 }, { "epoch": 0.16257473603867192, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.451727867126465, "learning_rate": 5.413310724883425e-07, "loss": 0.5258, "mean_token_accuracy": 0.8237208127975464, "num_tokens": 48744030.0, "step": 1278 }, { "epoch": 0.16270194631726242, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.405611515045166, "learning_rate": 5.417549809241205e-07, "loss": 0.5164, "mean_token_accuracy": 0.8370028138160706, "num_tokens": 48788401.0, "step": 1279 }, { "epoch": 0.16282915659585295, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.395648002624512, "learning_rate": 5.421788893598982e-07, "loss": 0.5204, "mean_token_accuracy": 0.8331254720687866, "num_tokens": 48823046.0, "step": 1280 }, { "epoch": 0.16295636687444345, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.420910358428955, "learning_rate": 5.42602797795676e-07, "loss": 0.5262, "mean_token_accuracy": 0.8301076292991638, "num_tokens": 48868293.0, "step": 1281 }, { "epoch": 0.16308357715303395, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.415699005126953, "learning_rate": 5.43026706231454e-07, "loss": 0.5544, "mean_token_accuracy": 0.821504533290863, "num_tokens": 48908775.0, "step": 1282 }, { "epoch": 0.16321078743162448, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.420016765594482, "learning_rate": 5.434506146672319e-07, "loss": 0.5317, "mean_token_accuracy": 0.8294187784194946, "num_tokens": 48957640.0, "step": 1283 }, { "epoch": 0.16333799771021498, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.485227584838867, "learning_rate": 5.438745231030097e-07, "loss": 0.5227, "mean_token_accuracy": 0.8288699388504028, "num_tokens": 48993601.0, "step": 1284 }, { "epoch": 0.16346520798880548, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.476283550262451, "learning_rate": 5.442984315387876e-07, "loss": 0.4292, "mean_token_accuracy": 0.8589116334915161, "num_tokens": 49028411.0, "step": 1285 }, { "epoch": 0.163592418267396, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.416102409362793, "learning_rate": 5.447223399745655e-07, "loss": 0.4929, "mean_token_accuracy": 0.8446089625358582, "num_tokens": 49067449.0, "step": 1286 }, { "epoch": 0.1637196285459865, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.480321407318115, "learning_rate": 5.451462484103433e-07, "loss": 0.4684, "mean_token_accuracy": 0.8493521809577942, "num_tokens": 49102650.0, "step": 1287 }, { "epoch": 0.163846838824577, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.500148296356201, "learning_rate": 5.455701568461212e-07, "loss": 0.4923, "mean_token_accuracy": 0.8392893671989441, "num_tokens": 49137110.0, "step": 1288 }, { "epoch": 0.16397404910316754, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.479267120361328, "learning_rate": 5.45994065281899e-07, "loss": 0.5632, "mean_token_accuracy": 0.8174782991409302, "num_tokens": 49177765.0, "step": 1289 }, { "epoch": 0.16410125938175804, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.543252468109131, "learning_rate": 5.46417973717677e-07, "loss": 0.5041, "mean_token_accuracy": 0.8349321484565735, "num_tokens": 49207860.0, "step": 1290 }, { "epoch": 0.16422846966034857, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.493131160736084, "learning_rate": 5.468418821534548e-07, "loss": 0.4369, "mean_token_accuracy": 0.8570700883865356, "num_tokens": 49239994.0, "step": 1291 }, { "epoch": 0.16435567993893907, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.460244178771973, "learning_rate": 5.472657905892327e-07, "loss": 0.4514, "mean_token_accuracy": 0.8543981313705444, "num_tokens": 49273464.0, "step": 1292 }, { "epoch": 0.16448289021752957, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.483189582824707, "learning_rate": 5.476896990250106e-07, "loss": 0.5649, "mean_token_accuracy": 0.8197824358940125, "num_tokens": 49313144.0, "step": 1293 }, { "epoch": 0.1646101004961201, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.413743019104004, "learning_rate": 5.481136074607885e-07, "loss": 0.4517, "mean_token_accuracy": 0.8487017750740051, "num_tokens": 49351392.0, "step": 1294 }, { "epoch": 0.1647373107747106, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.399075984954834, "learning_rate": 5.485375158965663e-07, "loss": 0.4537, "mean_token_accuracy": 0.8529237508773804, "num_tokens": 49396726.0, "step": 1295 }, { "epoch": 0.1648645210533011, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.425289154052734, "learning_rate": 5.489614243323442e-07, "loss": 0.4873, "mean_token_accuracy": 0.8417930006980896, "num_tokens": 49435856.0, "step": 1296 }, { "epoch": 0.16499173133189163, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.4703521728515625, "learning_rate": 5.49385332768122e-07, "loss": 0.4284, "mean_token_accuracy": 0.8585643172264099, "num_tokens": 49473695.0, "step": 1297 }, { "epoch": 0.16511894161048213, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.503960132598877, "learning_rate": 5.498092412039e-07, "loss": 0.5544, "mean_token_accuracy": 0.8271356821060181, "num_tokens": 49509165.0, "step": 1298 }, { "epoch": 0.16524615188907263, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.453623294830322, "learning_rate": 5.502331496396778e-07, "loss": 0.4586, "mean_token_accuracy": 0.8501558303833008, "num_tokens": 49553790.0, "step": 1299 }, { "epoch": 0.16537336216766316, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.450029373168945, "learning_rate": 5.506570580754557e-07, "loss": 0.4822, "mean_token_accuracy": 0.8437418937683105, "num_tokens": 49593124.0, "step": 1300 }, { "epoch": 0.16550057244625366, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.565706253051758, "learning_rate": 5.510809665112336e-07, "loss": 0.4698, "mean_token_accuracy": 0.8508696556091309, "num_tokens": 49632016.0, "step": 1301 }, { "epoch": 0.16562778272484416, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.460226058959961, "learning_rate": 5.515048749470113e-07, "loss": 0.4847, "mean_token_accuracy": 0.8447613716125488, "num_tokens": 49670689.0, "step": 1302 }, { "epoch": 0.1657549930034347, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.488092422485352, "learning_rate": 5.519287833827893e-07, "loss": 0.4705, "mean_token_accuracy": 0.8462695479393005, "num_tokens": 49709230.0, "step": 1303 }, { "epoch": 0.1658822032820252, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.540527820587158, "learning_rate": 5.523526918185671e-07, "loss": 0.526, "mean_token_accuracy": 0.8296504616737366, "num_tokens": 49741648.0, "step": 1304 }, { "epoch": 0.1660094135606157, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.4689154624938965, "learning_rate": 5.52776600254345e-07, "loss": 0.445, "mean_token_accuracy": 0.8510756492614746, "num_tokens": 49773906.0, "step": 1305 }, { "epoch": 0.16613662383920622, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.462726593017578, "learning_rate": 5.532005086901229e-07, "loss": 0.5041, "mean_token_accuracy": 0.842150092124939, "num_tokens": 49812645.0, "step": 1306 }, { "epoch": 0.16626383411779672, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.428126811981201, "learning_rate": 5.536244171259008e-07, "loss": 0.4731, "mean_token_accuracy": 0.8480358123779297, "num_tokens": 49854500.0, "step": 1307 }, { "epoch": 0.16639104439638722, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.5337371826171875, "learning_rate": 5.540483255616786e-07, "loss": 0.5667, "mean_token_accuracy": 0.8202627897262573, "num_tokens": 49889408.0, "step": 1308 }, { "epoch": 0.16651825467497774, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1682510375976562e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.479065895080566, "learning_rate": 5.544722339974566e-07, "loss": 0.4895, "mean_token_accuracy": 0.8426188826560974, "num_tokens": 49923581.0, "step": 1309 }, { "epoch": 0.16664546495356825, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1682510375976562e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.483667373657227, "learning_rate": 5.548961424332343e-07, "loss": 0.4782, "mean_token_accuracy": 0.8432028293609619, "num_tokens": 49959433.0, "step": 1310 }, { "epoch": 0.16677267523215875, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1682510375976562e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.500479698181152, "learning_rate": 5.553200508690123e-07, "loss": 0.4467, "mean_token_accuracy": 0.8562887907028198, "num_tokens": 50000307.0, "step": 1311 }, { "epoch": 0.16689988551074927, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1682510375976562e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.4922966957092285, "learning_rate": 5.557439593047901e-07, "loss": 0.5396, "mean_token_accuracy": 0.8256490230560303, "num_tokens": 50035587.0, "step": 1312 }, { "epoch": 0.16702709578933977, "ewc_loss": 0.006011962890625, "ewc_loss_diag": 1.1682510375976562e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.475286960601807, "learning_rate": 5.56167867740568e-07, "loss": 0.4778, "mean_token_accuracy": 0.8451712131500244, "num_tokens": 50079425.0, "step": 1313 }, { "epoch": 0.16715430606793028, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 6.50236701965332, "learning_rate": 5.565917761763459e-07, "loss": 0.4609, "mean_token_accuracy": 0.850495457649231, "num_tokens": 50110674.0, "step": 1314 }, { "epoch": 0.1672815163465208, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.559416770935059, "learning_rate": 5.570156846121238e-07, "loss": 0.4997, "mean_token_accuracy": 0.8378167748451233, "num_tokens": 50149033.0, "step": 1315 }, { "epoch": 0.1674087266251113, "ewc_loss": 0.006011962890625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 6.505650520324707, "learning_rate": 5.574395930479016e-07, "loss": 0.5197, "mean_token_accuracy": 0.8331392407417297, "num_tokens": 50185728.0, "step": 1316 }, { "epoch": 0.16753593690370183, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.523359298706055, "learning_rate": 5.578635014836796e-07, "loss": 0.4698, "mean_token_accuracy": 0.8477659225463867, "num_tokens": 50219671.0, "step": 1317 }, { "epoch": 0.16766314718229233, "ewc_loss": 0.006011962890625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 6.573551654815674, "learning_rate": 5.582874099194573e-07, "loss": 0.5289, "mean_token_accuracy": 0.8339203000068665, "num_tokens": 50251989.0, "step": 1318 }, { "epoch": 0.16779035746088283, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 6.531918048858643, "learning_rate": 5.587113183552353e-07, "loss": 0.5143, "mean_token_accuracy": 0.8351095914840698, "num_tokens": 50291079.0, "step": 1319 }, { "epoch": 0.16791756773947336, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.475699424743652, "learning_rate": 5.591352267910131e-07, "loss": 0.4705, "mean_token_accuracy": 0.8474352359771729, "num_tokens": 50332262.0, "step": 1320 }, { "epoch": 0.16804477801806386, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.551191329956055, "learning_rate": 5.59559135226791e-07, "loss": 0.5002, "mean_token_accuracy": 0.8417050242424011, "num_tokens": 50371278.0, "step": 1321 }, { "epoch": 0.16817198829665436, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.547199726104736, "learning_rate": 5.599830436625689e-07, "loss": 0.4776, "mean_token_accuracy": 0.846663773059845, "num_tokens": 50409463.0, "step": 1322 }, { "epoch": 0.1682991985752449, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.5086541175842285, "learning_rate": 5.604069520983468e-07, "loss": 0.4716, "mean_token_accuracy": 0.8487245440483093, "num_tokens": 50446745.0, "step": 1323 }, { "epoch": 0.1684264088538354, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.60550594329834, "learning_rate": 5.608308605341246e-07, "loss": 0.4636, "mean_token_accuracy": 0.8477699160575867, "num_tokens": 50479725.0, "step": 1324 }, { "epoch": 0.1685536191324259, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.569990158081055, "learning_rate": 5.612547689699024e-07, "loss": 0.5263, "mean_token_accuracy": 0.8314695954322815, "num_tokens": 50514734.0, "step": 1325 }, { "epoch": 0.16868082941101642, "ewc_loss": 0.006011962890625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 6.52599573135376, "learning_rate": 5.616786774056803e-07, "loss": 0.5046, "mean_token_accuracy": 0.8403463959693909, "num_tokens": 50555165.0, "step": 1326 }, { "epoch": 0.16880803968960692, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.503814220428467, "learning_rate": 5.621025858414582e-07, "loss": 0.5059, "mean_token_accuracy": 0.8368822336196899, "num_tokens": 50598251.0, "step": 1327 }, { "epoch": 0.16893524996819742, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.503130912780762, "learning_rate": 5.625264942772361e-07, "loss": 0.4977, "mean_token_accuracy": 0.8416271805763245, "num_tokens": 50638423.0, "step": 1328 }, { "epoch": 0.16906246024678795, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.53454065322876, "learning_rate": 5.629504027130139e-07, "loss": 0.5556, "mean_token_accuracy": 0.8305686116218567, "num_tokens": 50678226.0, "step": 1329 }, { "epoch": 0.16918967052537845, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.567451477050781, "learning_rate": 5.633743111487919e-07, "loss": 0.556, "mean_token_accuracy": 0.8198763132095337, "num_tokens": 50717542.0, "step": 1330 }, { "epoch": 0.16931688080396895, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.576108932495117, "learning_rate": 5.637982195845697e-07, "loss": 0.4367, "mean_token_accuracy": 0.8574879169464111, "num_tokens": 50752914.0, "step": 1331 }, { "epoch": 0.16944409108255948, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.491242408752441, "learning_rate": 5.642221280203476e-07, "loss": 0.4357, "mean_token_accuracy": 0.8595606088638306, "num_tokens": 50789276.0, "step": 1332 }, { "epoch": 0.16957130136114998, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.675495147705078, "learning_rate": 5.646460364561254e-07, "loss": 0.456, "mean_token_accuracy": 0.8501638174057007, "num_tokens": 50820759.0, "step": 1333 }, { "epoch": 0.16969851163974048, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.571177959442139, "learning_rate": 5.650699448919033e-07, "loss": 0.5467, "mean_token_accuracy": 0.8255022764205933, "num_tokens": 50863241.0, "step": 1334 }, { "epoch": 0.169825721918331, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.530088901519775, "learning_rate": 5.654938533276812e-07, "loss": 0.4855, "mean_token_accuracy": 0.8421151638031006, "num_tokens": 50904296.0, "step": 1335 }, { "epoch": 0.1699529321969215, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.586195468902588, "learning_rate": 5.659177617634591e-07, "loss": 0.4722, "mean_token_accuracy": 0.8480689525604248, "num_tokens": 50942180.0, "step": 1336 }, { "epoch": 0.170080142475512, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.532257556915283, "learning_rate": 5.663416701992369e-07, "loss": 0.4982, "mean_token_accuracy": 0.8411810398101807, "num_tokens": 50980287.0, "step": 1337 }, { "epoch": 0.17020735275410254, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.643066883087158, "learning_rate": 5.667655786350149e-07, "loss": 0.5604, "mean_token_accuracy": 0.8145418167114258, "num_tokens": 51014218.0, "step": 1338 }, { "epoch": 0.17033456303269304, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.545076847076416, "learning_rate": 5.671894870707927e-07, "loss": 0.4782, "mean_token_accuracy": 0.8480405211448669, "num_tokens": 51055915.0, "step": 1339 }, { "epoch": 0.17046177331128357, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.551835060119629, "learning_rate": 5.676133955065705e-07, "loss": 0.4221, "mean_token_accuracy": 0.8598020076751709, "num_tokens": 51090684.0, "step": 1340 }, { "epoch": 0.17058898358987407, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.566898345947266, "learning_rate": 5.680373039423484e-07, "loss": 0.468, "mean_token_accuracy": 0.8471733927726746, "num_tokens": 51130156.0, "step": 1341 }, { "epoch": 0.17071619386846457, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.599681854248047, "learning_rate": 5.684612123781263e-07, "loss": 0.5085, "mean_token_accuracy": 0.8349013328552246, "num_tokens": 51171282.0, "step": 1342 }, { "epoch": 0.1708434041470551, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.6086554527282715, "learning_rate": 5.688851208139042e-07, "loss": 0.5245, "mean_token_accuracy": 0.8299140930175781, "num_tokens": 51210616.0, "step": 1343 }, { "epoch": 0.1709706144256456, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.578537940979004, "learning_rate": 5.69309029249682e-07, "loss": 0.5103, "mean_token_accuracy": 0.8403749465942383, "num_tokens": 51253357.0, "step": 1344 }, { "epoch": 0.1710978247042361, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.569891452789307, "learning_rate": 5.697329376854599e-07, "loss": 0.4903, "mean_token_accuracy": 0.8409742116928101, "num_tokens": 51292671.0, "step": 1345 }, { "epoch": 0.17122503498282662, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.607940196990967, "learning_rate": 5.701568461212378e-07, "loss": 0.4304, "mean_token_accuracy": 0.8604565858840942, "num_tokens": 51329145.0, "step": 1346 }, { "epoch": 0.17135224526141712, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.59044885635376, "learning_rate": 5.705807545570157e-07, "loss": 0.4396, "mean_token_accuracy": 0.8557349443435669, "num_tokens": 51371265.0, "step": 1347 }, { "epoch": 0.17147945554000762, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.619142055511475, "learning_rate": 5.710046629927934e-07, "loss": 0.4762, "mean_token_accuracy": 0.844785749912262, "num_tokens": 51410262.0, "step": 1348 }, { "epoch": 0.17160666581859815, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.603003025054932, "learning_rate": 5.714285714285714e-07, "loss": 0.447, "mean_token_accuracy": 0.8551160097122192, "num_tokens": 51447796.0, "step": 1349 }, { "epoch": 0.17173387609718865, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.7186360359191895, "learning_rate": 5.718524798643492e-07, "loss": 0.511, "mean_token_accuracy": 0.8377329111099243, "num_tokens": 51482677.0, "step": 1350 }, { "epoch": 0.17186108637577915, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.588460922241211, "learning_rate": 5.722763883001272e-07, "loss": 0.512, "mean_token_accuracy": 0.8373594284057617, "num_tokens": 51524111.0, "step": 1351 }, { "epoch": 0.17198829665436968, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.600662708282471, "learning_rate": 5.72700296735905e-07, "loss": 0.4846, "mean_token_accuracy": 0.8426946401596069, "num_tokens": 51561008.0, "step": 1352 }, { "epoch": 0.17211550693296018, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.64558744430542, "learning_rate": 5.731242051716829e-07, "loss": 0.5384, "mean_token_accuracy": 0.8247929215431213, "num_tokens": 51603091.0, "step": 1353 }, { "epoch": 0.17224271721155068, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.609704971313477, "learning_rate": 5.735481136074608e-07, "loss": 0.454, "mean_token_accuracy": 0.8503435850143433, "num_tokens": 51637789.0, "step": 1354 }, { "epoch": 0.1723699274901412, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.674647331237793, "learning_rate": 5.739720220432386e-07, "loss": 0.4583, "mean_token_accuracy": 0.8487979173660278, "num_tokens": 51674207.0, "step": 1355 }, { "epoch": 0.1724971377687317, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.620929718017578, "learning_rate": 5.743959304790164e-07, "loss": 0.5537, "mean_token_accuracy": 0.825089693069458, "num_tokens": 51715109.0, "step": 1356 }, { "epoch": 0.1726243480473222, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.65313720703125, "learning_rate": 5.748198389147944e-07, "loss": 0.5014, "mean_token_accuracy": 0.8408308625221252, "num_tokens": 51753373.0, "step": 1357 }, { "epoch": 0.17275155832591274, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.654414653778076, "learning_rate": 5.752437473505722e-07, "loss": 0.5386, "mean_token_accuracy": 0.830847978591919, "num_tokens": 51795100.0, "step": 1358 }, { "epoch": 0.17287876860450324, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.6594953536987305, "learning_rate": 5.756676557863502e-07, "loss": 0.5124, "mean_token_accuracy": 0.8345775604248047, "num_tokens": 51831840.0, "step": 1359 }, { "epoch": 0.17300597888309374, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.595643997192383, "learning_rate": 5.76091564222128e-07, "loss": 0.4647, "mean_token_accuracy": 0.8500387072563171, "num_tokens": 51870227.0, "step": 1360 }, { "epoch": 0.17313318916168427, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.678126335144043, "learning_rate": 5.765154726579059e-07, "loss": 0.4902, "mean_token_accuracy": 0.8426499962806702, "num_tokens": 51908251.0, "step": 1361 }, { "epoch": 0.17326039944027477, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.650285243988037, "learning_rate": 5.769393810936838e-07, "loss": 0.4803, "mean_token_accuracy": 0.8442105650901794, "num_tokens": 51944531.0, "step": 1362 }, { "epoch": 0.17338760971886527, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.617868423461914, "learning_rate": 5.773632895294616e-07, "loss": 0.5333, "mean_token_accuracy": 0.8274697065353394, "num_tokens": 51985742.0, "step": 1363 }, { "epoch": 0.1735148199974558, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.674892425537109, "learning_rate": 5.777871979652394e-07, "loss": 0.4711, "mean_token_accuracy": 0.8478685617446899, "num_tokens": 52022036.0, "step": 1364 }, { "epoch": 0.1736420302760463, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.621866703033447, "learning_rate": 5.782111064010173e-07, "loss": 0.4943, "mean_token_accuracy": 0.8397983908653259, "num_tokens": 52066077.0, "step": 1365 }, { "epoch": 0.17376924055463683, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.607933521270752, "learning_rate": 5.786350148367952e-07, "loss": 0.4853, "mean_token_accuracy": 0.8430113792419434, "num_tokens": 52111280.0, "step": 1366 }, { "epoch": 0.17389645083322733, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.637236595153809, "learning_rate": 5.790589232725731e-07, "loss": 0.4838, "mean_token_accuracy": 0.8417676687240601, "num_tokens": 52153148.0, "step": 1367 }, { "epoch": 0.17402366111181783, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.674982070922852, "learning_rate": 5.79482831708351e-07, "loss": 0.4685, "mean_token_accuracy": 0.8479984998703003, "num_tokens": 52186722.0, "step": 1368 }, { "epoch": 0.17415087139040836, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.730311870574951, "learning_rate": 5.799067401441288e-07, "loss": 0.478, "mean_token_accuracy": 0.843091607093811, "num_tokens": 52220837.0, "step": 1369 }, { "epoch": 0.17427808166899886, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.679661273956299, "learning_rate": 5.803306485799068e-07, "loss": 0.4629, "mean_token_accuracy": 0.8512871265411377, "num_tokens": 52257049.0, "step": 1370 }, { "epoch": 0.17440529194758936, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.575636863708496, "learning_rate": 5.807545570156845e-07, "loss": 0.4402, "mean_token_accuracy": 0.8551344871520996, "num_tokens": 52297632.0, "step": 1371 }, { "epoch": 0.1745325022261799, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.766218662261963, "learning_rate": 5.811784654514624e-07, "loss": 0.5386, "mean_token_accuracy": 0.8274873495101929, "num_tokens": 52331008.0, "step": 1372 }, { "epoch": 0.1746597125047704, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.656485557556152, "learning_rate": 5.816023738872403e-07, "loss": 0.4703, "mean_token_accuracy": 0.8489055633544922, "num_tokens": 52369420.0, "step": 1373 }, { "epoch": 0.1747869227833609, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.6574835777282715, "learning_rate": 5.820262823230182e-07, "loss": 0.5003, "mean_token_accuracy": 0.8388350009918213, "num_tokens": 52411759.0, "step": 1374 }, { "epoch": 0.17491413306195142, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.679786205291748, "learning_rate": 5.824501907587961e-07, "loss": 0.4615, "mean_token_accuracy": 0.8498883843421936, "num_tokens": 52446459.0, "step": 1375 }, { "epoch": 0.17504134334054192, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.720802307128906, "learning_rate": 5.82874099194574e-07, "loss": 0.5444, "mean_token_accuracy": 0.8237522840499878, "num_tokens": 52484615.0, "step": 1376 }, { "epoch": 0.17516855361913242, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.650234222412109, "learning_rate": 5.832980076303518e-07, "loss": 0.5552, "mean_token_accuracy": 0.8215366005897522, "num_tokens": 52525195.0, "step": 1377 }, { "epoch": 0.17529576389772294, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.688289642333984, "learning_rate": 5.837219160661297e-07, "loss": 0.4396, "mean_token_accuracy": 0.8555231094360352, "num_tokens": 52558783.0, "step": 1378 }, { "epoch": 0.17542297417631345, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.691647052764893, "learning_rate": 5.841458245019075e-07, "loss": 0.4345, "mean_token_accuracy": 0.8572288751602173, "num_tokens": 52599176.0, "step": 1379 }, { "epoch": 0.17555018445490395, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.740603923797607, "learning_rate": 5.845697329376855e-07, "loss": 0.491, "mean_token_accuracy": 0.8450087308883667, "num_tokens": 52630865.0, "step": 1380 }, { "epoch": 0.17567739473349447, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.697409629821777, "learning_rate": 5.849936413734633e-07, "loss": 0.5426, "mean_token_accuracy": 0.8257197141647339, "num_tokens": 52671702.0, "step": 1381 }, { "epoch": 0.17580460501208497, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.651974678039551, "learning_rate": 5.854175498092412e-07, "loss": 0.4472, "mean_token_accuracy": 0.8547462224960327, "num_tokens": 52712288.0, "step": 1382 }, { "epoch": 0.17593181529067548, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.678081035614014, "learning_rate": 5.858414582450191e-07, "loss": 0.4898, "mean_token_accuracy": 0.8429662585258484, "num_tokens": 52756569.0, "step": 1383 }, { "epoch": 0.176059025569266, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.684578895568848, "learning_rate": 5.86265366680797e-07, "loss": 0.5126, "mean_token_accuracy": 0.8365252017974854, "num_tokens": 52796195.0, "step": 1384 }, { "epoch": 0.1761862358478565, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.687442302703857, "learning_rate": 5.866892751165748e-07, "loss": 0.4832, "mean_token_accuracy": 0.8453660011291504, "num_tokens": 52839687.0, "step": 1385 }, { "epoch": 0.176313446126447, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.730164527893066, "learning_rate": 5.871131835523526e-07, "loss": 0.5379, "mean_token_accuracy": 0.8279509544372559, "num_tokens": 52875660.0, "step": 1386 }, { "epoch": 0.17644065640503753, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.700268268585205, "learning_rate": 5.875370919881305e-07, "loss": 0.4941, "mean_token_accuracy": 0.8389169573783875, "num_tokens": 52921220.0, "step": 1387 }, { "epoch": 0.17656786668362803, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.691385269165039, "learning_rate": 5.879610004239084e-07, "loss": 0.5476, "mean_token_accuracy": 0.8280310034751892, "num_tokens": 52960042.0, "step": 1388 }, { "epoch": 0.17669507696221853, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.746585845947266, "learning_rate": 5.883849088596863e-07, "loss": 0.5319, "mean_token_accuracy": 0.8312594890594482, "num_tokens": 53002817.0, "step": 1389 }, { "epoch": 0.17682228724080906, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.737503528594971, "learning_rate": 5.888088172954641e-07, "loss": 0.5051, "mean_token_accuracy": 0.8381063938140869, "num_tokens": 53037751.0, "step": 1390 }, { "epoch": 0.17694949751939956, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.720830917358398, "learning_rate": 5.892327257312421e-07, "loss": 0.4184, "mean_token_accuracy": 0.8651413917541504, "num_tokens": 53074286.0, "step": 1391 }, { "epoch": 0.1770767077979901, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.686460018157959, "learning_rate": 5.896566341670199e-07, "loss": 0.4401, "mean_token_accuracy": 0.8519959449768066, "num_tokens": 53113525.0, "step": 1392 }, { "epoch": 0.1772039180765806, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.728853702545166, "learning_rate": 5.900805426027977e-07, "loss": 0.4831, "mean_token_accuracy": 0.8430871963500977, "num_tokens": 53150137.0, "step": 1393 }, { "epoch": 0.1773311283551711, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.688074588775635, "learning_rate": 5.905044510385756e-07, "loss": 0.4361, "mean_token_accuracy": 0.858153760433197, "num_tokens": 53189519.0, "step": 1394 }, { "epoch": 0.17745833863376162, "ewc_loss": 0.006439208984375, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.762232780456543, "learning_rate": 5.909283594743535e-07, "loss": 0.5469, "mean_token_accuracy": 0.8245626091957092, "num_tokens": 53227452.0, "step": 1395 }, { "epoch": 0.17758554891235212, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.715723514556885, "learning_rate": 5.913522679101314e-07, "loss": 0.4789, "mean_token_accuracy": 0.8414599895477295, "num_tokens": 53265875.0, "step": 1396 }, { "epoch": 0.17771275919094262, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.7500386238098145, "learning_rate": 5.917761763459093e-07, "loss": 0.5392, "mean_token_accuracy": 0.8263326287269592, "num_tokens": 53305477.0, "step": 1397 }, { "epoch": 0.17783996946953315, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.727284908294678, "learning_rate": 5.922000847816871e-07, "loss": 0.459, "mean_token_accuracy": 0.8522096872329712, "num_tokens": 53342800.0, "step": 1398 }, { "epoch": 0.17796717974812365, "ewc_loss": 0.006439208984375, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.715970993041992, "learning_rate": 5.926239932174651e-07, "loss": 0.5236, "mean_token_accuracy": 0.8322756290435791, "num_tokens": 53387476.0, "step": 1399 }, { "epoch": 0.17809439002671415, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.709547519683838, "learning_rate": 5.930479016532429e-07, "loss": 0.4802, "mean_token_accuracy": 0.8443424701690674, "num_tokens": 53424628.0, "step": 1400 }, { "epoch": 0.17822160030530468, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.70196533203125, "learning_rate": 5.934718100890207e-07, "loss": 0.4613, "mean_token_accuracy": 0.8476588726043701, "num_tokens": 53464459.0, "step": 1401 }, { "epoch": 0.17834881058389518, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.812124729156494, "learning_rate": 5.938957185247986e-07, "loss": 0.5099, "mean_token_accuracy": 0.8330515027046204, "num_tokens": 53501721.0, "step": 1402 }, { "epoch": 0.17847602086248568, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.760995388031006, "learning_rate": 5.943196269605765e-07, "loss": 0.4707, "mean_token_accuracy": 0.8435425758361816, "num_tokens": 53536607.0, "step": 1403 }, { "epoch": 0.1786032311410762, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.749380111694336, "learning_rate": 5.947435353963544e-07, "loss": 0.5214, "mean_token_accuracy": 0.8364429473876953, "num_tokens": 53569809.0, "step": 1404 }, { "epoch": 0.1787304414196667, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.7946367263793945, "learning_rate": 5.951674438321323e-07, "loss": 0.4981, "mean_token_accuracy": 0.8398551940917969, "num_tokens": 53604423.0, "step": 1405 }, { "epoch": 0.1788576516982572, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.775123596191406, "learning_rate": 5.955913522679101e-07, "loss": 0.4943, "mean_token_accuracy": 0.8366378545761108, "num_tokens": 53642145.0, "step": 1406 }, { "epoch": 0.17898486197684774, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.725897789001465, "learning_rate": 5.96015260703688e-07, "loss": 0.5692, "mean_token_accuracy": 0.8166361451148987, "num_tokens": 53682998.0, "step": 1407 }, { "epoch": 0.17911207225543824, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.750565528869629, "learning_rate": 5.964391691394659e-07, "loss": 0.502, "mean_token_accuracy": 0.8403767347335815, "num_tokens": 53729764.0, "step": 1408 }, { "epoch": 0.17923928253402874, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.793202877044678, "learning_rate": 5.968630775752436e-07, "loss": 0.5357, "mean_token_accuracy": 0.8287598490715027, "num_tokens": 53764522.0, "step": 1409 }, { "epoch": 0.17936649281261927, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.821805477142334, "learning_rate": 5.972869860110216e-07, "loss": 0.5174, "mean_token_accuracy": 0.8302526473999023, "num_tokens": 53799791.0, "step": 1410 }, { "epoch": 0.17949370309120977, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.776648044586182, "learning_rate": 5.977108944467994e-07, "loss": 0.5559, "mean_token_accuracy": 0.826404333114624, "num_tokens": 53840181.0, "step": 1411 }, { "epoch": 0.17962091336980027, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.761368274688721, "learning_rate": 5.981348028825774e-07, "loss": 0.4625, "mean_token_accuracy": 0.852476954460144, "num_tokens": 53882813.0, "step": 1412 }, { "epoch": 0.1797481236483908, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.765516757965088, "learning_rate": 5.985587113183552e-07, "loss": 0.5105, "mean_token_accuracy": 0.8364595770835876, "num_tokens": 53922198.0, "step": 1413 }, { "epoch": 0.1798753339269813, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.838332653045654, "learning_rate": 5.989826197541331e-07, "loss": 0.5056, "mean_token_accuracy": 0.8374490737915039, "num_tokens": 53959390.0, "step": 1414 }, { "epoch": 0.18000254420557182, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.728063106536865, "learning_rate": 5.99406528189911e-07, "loss": 0.533, "mean_token_accuracy": 0.8341996073722839, "num_tokens": 54005707.0, "step": 1415 }, { "epoch": 0.18012975448416232, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.78354549407959, "learning_rate": 5.998304366256888e-07, "loss": 0.4975, "mean_token_accuracy": 0.8391450047492981, "num_tokens": 54047808.0, "step": 1416 }, { "epoch": 0.18025696476275282, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.794656276702881, "learning_rate": 6.002543450614666e-07, "loss": 0.4222, "mean_token_accuracy": 0.8651931881904602, "num_tokens": 54082929.0, "step": 1417 }, { "epoch": 0.18038417504134335, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.810007095336914, "learning_rate": 6.006782534972446e-07, "loss": 0.483, "mean_token_accuracy": 0.8442813158035278, "num_tokens": 54120199.0, "step": 1418 }, { "epoch": 0.18051138531993385, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.803928375244141, "learning_rate": 6.011021619330224e-07, "loss": 0.5187, "mean_token_accuracy": 0.8317165374755859, "num_tokens": 54158629.0, "step": 1419 }, { "epoch": 0.18063859559852435, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.836326599121094, "learning_rate": 6.015260703688004e-07, "loss": 0.4982, "mean_token_accuracy": 0.8365097641944885, "num_tokens": 54194772.0, "step": 1420 }, { "epoch": 0.18076580587711488, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.83160924911499, "learning_rate": 6.019499788045782e-07, "loss": 0.5059, "mean_token_accuracy": 0.837965726852417, "num_tokens": 54229825.0, "step": 1421 }, { "epoch": 0.18089301615570538, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.845119953155518, "learning_rate": 6.023738872403561e-07, "loss": 0.462, "mean_token_accuracy": 0.8491861820220947, "num_tokens": 54260051.0, "step": 1422 }, { "epoch": 0.18102022643429588, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.808323860168457, "learning_rate": 6.02797795676134e-07, "loss": 0.4888, "mean_token_accuracy": 0.8463253974914551, "num_tokens": 54297865.0, "step": 1423 }, { "epoch": 0.1811474367128864, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.829547882080078, "learning_rate": 6.032217041119118e-07, "loss": 0.5033, "mean_token_accuracy": 0.8330444097518921, "num_tokens": 54334618.0, "step": 1424 }, { "epoch": 0.1812746469914769, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.799554824829102, "learning_rate": 6.036456125476896e-07, "loss": 0.5398, "mean_token_accuracy": 0.8253192901611328, "num_tokens": 54369761.0, "step": 1425 }, { "epoch": 0.1814018572700674, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.786143779754639, "learning_rate": 6.040695209834675e-07, "loss": 0.4396, "mean_token_accuracy": 0.8547921180725098, "num_tokens": 54411549.0, "step": 1426 }, { "epoch": 0.18152906754865794, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.813441276550293, "learning_rate": 6.044934294192454e-07, "loss": 0.5247, "mean_token_accuracy": 0.8321584463119507, "num_tokens": 54449137.0, "step": 1427 }, { "epoch": 0.18165627782724844, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.764285087585449, "learning_rate": 6.049173378550233e-07, "loss": 0.4797, "mean_token_accuracy": 0.8467152118682861, "num_tokens": 54494604.0, "step": 1428 }, { "epoch": 0.18178348810583894, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.75718879699707, "learning_rate": 6.053412462908012e-07, "loss": 0.46, "mean_token_accuracy": 0.8490443825721741, "num_tokens": 54533533.0, "step": 1429 }, { "epoch": 0.18191069838442947, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.801318645477295, "learning_rate": 6.05765154726579e-07, "loss": 0.5323, "mean_token_accuracy": 0.8308807611465454, "num_tokens": 54575211.0, "step": 1430 }, { "epoch": 0.18203790866301997, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.821334362030029, "learning_rate": 6.061890631623569e-07, "loss": 0.4383, "mean_token_accuracy": 0.856891393661499, "num_tokens": 54618308.0, "step": 1431 }, { "epoch": 0.18216511894161047, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.776559829711914, "learning_rate": 6.066129715981347e-07, "loss": 0.4769, "mean_token_accuracy": 0.844810962677002, "num_tokens": 54660260.0, "step": 1432 }, { "epoch": 0.182292329220201, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.896547317504883, "learning_rate": 6.070368800339126e-07, "loss": 0.448, "mean_token_accuracy": 0.8532437086105347, "num_tokens": 54692056.0, "step": 1433 }, { "epoch": 0.1824195394987915, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.807809829711914, "learning_rate": 6.074607884696905e-07, "loss": 0.4696, "mean_token_accuracy": 0.8458710312843323, "num_tokens": 54729252.0, "step": 1434 }, { "epoch": 0.182546749777382, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.8926544189453125, "learning_rate": 6.078846969054684e-07, "loss": 0.4927, "mean_token_accuracy": 0.8392057418823242, "num_tokens": 54763422.0, "step": 1435 }, { "epoch": 0.18267396005597253, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.847349166870117, "learning_rate": 6.083086053412463e-07, "loss": 0.4686, "mean_token_accuracy": 0.8480095863342285, "num_tokens": 54803494.0, "step": 1436 }, { "epoch": 0.18280117033456303, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.86088752746582, "learning_rate": 6.087325137770242e-07, "loss": 0.4573, "mean_token_accuracy": 0.8502137064933777, "num_tokens": 54839349.0, "step": 1437 }, { "epoch": 0.18292838061315353, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.829499244689941, "learning_rate": 6.09156422212802e-07, "loss": 0.5308, "mean_token_accuracy": 0.833531379699707, "num_tokens": 54881941.0, "step": 1438 }, { "epoch": 0.18305559089174406, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.87271785736084, "learning_rate": 6.095803306485799e-07, "loss": 0.4815, "mean_token_accuracy": 0.8417752981185913, "num_tokens": 54922927.0, "step": 1439 }, { "epoch": 0.18318280117033456, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.833846092224121, "learning_rate": 6.100042390843577e-07, "loss": 0.5086, "mean_token_accuracy": 0.8353359699249268, "num_tokens": 54963885.0, "step": 1440 }, { "epoch": 0.1833100114489251, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.794337749481201, "learning_rate": 6.104281475201356e-07, "loss": 0.459, "mean_token_accuracy": 0.8492467403411865, "num_tokens": 55001284.0, "step": 1441 }, { "epoch": 0.1834372217275156, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.856637954711914, "learning_rate": 6.108520559559135e-07, "loss": 0.5184, "mean_token_accuracy": 0.833710789680481, "num_tokens": 55034813.0, "step": 1442 }, { "epoch": 0.1835644320061061, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.813024520874023, "learning_rate": 6.112759643916914e-07, "loss": 0.4244, "mean_token_accuracy": 0.8613380789756775, "num_tokens": 55070620.0, "step": 1443 }, { "epoch": 0.18369164228469662, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.89747953414917, "learning_rate": 6.116998728274693e-07, "loss": 0.506, "mean_token_accuracy": 0.8418519496917725, "num_tokens": 55108394.0, "step": 1444 }, { "epoch": 0.18381885256328712, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.875852108001709, "learning_rate": 6.121237812632472e-07, "loss": 0.5094, "mean_token_accuracy": 0.8342417478561401, "num_tokens": 55148170.0, "step": 1445 }, { "epoch": 0.18394606284187762, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.799069404602051, "learning_rate": 6.125476896990249e-07, "loss": 0.4518, "mean_token_accuracy": 0.8554152250289917, "num_tokens": 55188102.0, "step": 1446 }, { "epoch": 0.18407327312046814, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.831380367279053, "learning_rate": 6.129715981348028e-07, "loss": 0.4906, "mean_token_accuracy": 0.8456223607063293, "num_tokens": 55225893.0, "step": 1447 }, { "epoch": 0.18420048339905865, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.894814968109131, "learning_rate": 6.133955065705807e-07, "loss": 0.51, "mean_token_accuracy": 0.8330570459365845, "num_tokens": 55265321.0, "step": 1448 }, { "epoch": 0.18432769367764915, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.860189914703369, "learning_rate": 6.138194150063585e-07, "loss": 0.4709, "mean_token_accuracy": 0.8483940958976746, "num_tokens": 55304179.0, "step": 1449 }, { "epoch": 0.18445490395623967, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.835434913635254, "learning_rate": 6.142433234421365e-07, "loss": 0.5047, "mean_token_accuracy": 0.8362019062042236, "num_tokens": 55343266.0, "step": 1450 }, { "epoch": 0.18458211423483017, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.875003337860107, "learning_rate": 6.146672318779143e-07, "loss": 0.5079, "mean_token_accuracy": 0.8364130854606628, "num_tokens": 55381147.0, "step": 1451 }, { "epoch": 0.18470932451342068, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.896708965301514, "learning_rate": 6.150911403136923e-07, "loss": 0.5081, "mean_token_accuracy": 0.8383884429931641, "num_tokens": 55414855.0, "step": 1452 }, { "epoch": 0.1848365347920112, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.880168437957764, "learning_rate": 6.155150487494701e-07, "loss": 0.4324, "mean_token_accuracy": 0.8591920137405396, "num_tokens": 55449699.0, "step": 1453 }, { "epoch": 0.1849637450706017, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.834782600402832, "learning_rate": 6.159389571852479e-07, "loss": 0.4596, "mean_token_accuracy": 0.8526906967163086, "num_tokens": 55493202.0, "step": 1454 }, { "epoch": 0.1850909553491922, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.842560768127441, "learning_rate": 6.163628656210258e-07, "loss": 0.4807, "mean_token_accuracy": 0.8472057580947876, "num_tokens": 55539873.0, "step": 1455 }, { "epoch": 0.18521816562778273, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.929863452911377, "learning_rate": 6.167867740568037e-07, "loss": 0.541, "mean_token_accuracy": 0.8284560441970825, "num_tokens": 55576338.0, "step": 1456 }, { "epoch": 0.18534537590637323, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.832523345947266, "learning_rate": 6.172106824925815e-07, "loss": 0.4617, "mean_token_accuracy": 0.8484803438186646, "num_tokens": 55618330.0, "step": 1457 }, { "epoch": 0.18547258618496373, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.869147777557373, "learning_rate": 6.176345909283595e-07, "loss": 0.4788, "mean_token_accuracy": 0.8447317481040955, "num_tokens": 55657394.0, "step": 1458 }, { "epoch": 0.18559979646355426, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.890340328216553, "learning_rate": 6.180584993641373e-07, "loss": 0.4641, "mean_token_accuracy": 0.8493887186050415, "num_tokens": 55695087.0, "step": 1459 }, { "epoch": 0.18572700674214476, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.828192234039307, "learning_rate": 6.184824077999153e-07, "loss": 0.4485, "mean_token_accuracy": 0.8531022071838379, "num_tokens": 55730852.0, "step": 1460 }, { "epoch": 0.18585421702073526, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.918715476989746, "learning_rate": 6.189063162356931e-07, "loss": 0.545, "mean_token_accuracy": 0.8272437453269958, "num_tokens": 55765678.0, "step": 1461 }, { "epoch": 0.1859814272993258, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.939444541931152, "learning_rate": 6.193302246714709e-07, "loss": 0.471, "mean_token_accuracy": 0.8500447869300842, "num_tokens": 55804409.0, "step": 1462 }, { "epoch": 0.1861086375779163, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.9123735427856445, "learning_rate": 6.197541331072488e-07, "loss": 0.5, "mean_token_accuracy": 0.8377553224563599, "num_tokens": 55839520.0, "step": 1463 }, { "epoch": 0.1862358478565068, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.868248462677002, "learning_rate": 6.201780415430267e-07, "loss": 0.4798, "mean_token_accuracy": 0.8443479537963867, "num_tokens": 55879089.0, "step": 1464 }, { "epoch": 0.18636305813509732, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.934084892272949, "learning_rate": 6.206019499788045e-07, "loss": 0.5226, "mean_token_accuracy": 0.8321128487586975, "num_tokens": 55916655.0, "step": 1465 }, { "epoch": 0.18649026841368782, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.864934921264648, "learning_rate": 6.210258584145825e-07, "loss": 0.4582, "mean_token_accuracy": 0.8482651710510254, "num_tokens": 55953828.0, "step": 1466 }, { "epoch": 0.18661747869227835, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.893655776977539, "learning_rate": 6.214497668503603e-07, "loss": 0.507, "mean_token_accuracy": 0.8328455686569214, "num_tokens": 55992037.0, "step": 1467 }, { "epoch": 0.18674468897086885, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.970365524291992, "learning_rate": 6.218736752861383e-07, "loss": 0.4679, "mean_token_accuracy": 0.8500469923019409, "num_tokens": 56020424.0, "step": 1468 }, { "epoch": 0.18687189924945935, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 6.8978071212768555, "learning_rate": 6.22297583721916e-07, "loss": 0.5005, "mean_token_accuracy": 0.8436794877052307, "num_tokens": 56060839.0, "step": 1469 }, { "epoch": 0.18699910952804988, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.916677951812744, "learning_rate": 6.227214921576938e-07, "loss": 0.5299, "mean_token_accuracy": 0.8341292142868042, "num_tokens": 56099095.0, "step": 1470 }, { "epoch": 0.18712631980664038, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.918368339538574, "learning_rate": 6.231454005934718e-07, "loss": 0.4267, "mean_token_accuracy": 0.8613327741622925, "num_tokens": 56133805.0, "step": 1471 }, { "epoch": 0.18725353008523088, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.928464889526367, "learning_rate": 6.235693090292496e-07, "loss": 0.5454, "mean_token_accuracy": 0.828964114189148, "num_tokens": 56173064.0, "step": 1472 }, { "epoch": 0.1873807403638214, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 6.871424674987793, "learning_rate": 6.239932174650275e-07, "loss": 0.4278, "mean_token_accuracy": 0.8623827695846558, "num_tokens": 56213285.0, "step": 1473 }, { "epoch": 0.1875079506424119, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.977357387542725, "learning_rate": 6.244171259008054e-07, "loss": 0.5094, "mean_token_accuracy": 0.8329942226409912, "num_tokens": 56248218.0, "step": 1474 }, { "epoch": 0.1876351609210024, "ewc_loss": 0.006927490234375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 6.916926383972168, "learning_rate": 6.248410343365833e-07, "loss": 0.4539, "mean_token_accuracy": 0.8527426719665527, "num_tokens": 56286098.0, "step": 1475 }, { "epoch": 0.18776237119959294, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.893306255340576, "learning_rate": 6.252649427723612e-07, "loss": 0.4618, "mean_token_accuracy": 0.8519560098648071, "num_tokens": 56329732.0, "step": 1476 }, { "epoch": 0.18788958147818344, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.938730716705322, "learning_rate": 6.25688851208139e-07, "loss": 0.4661, "mean_token_accuracy": 0.8505332469940186, "num_tokens": 56367838.0, "step": 1477 }, { "epoch": 0.18801679175677394, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.9599127769470215, "learning_rate": 6.261127596439168e-07, "loss": 0.5192, "mean_token_accuracy": 0.8319861888885498, "num_tokens": 56413059.0, "step": 1478 }, { "epoch": 0.18814400203536447, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.928142547607422, "learning_rate": 6.265366680796948e-07, "loss": 0.4219, "mean_token_accuracy": 0.857699990272522, "num_tokens": 56453806.0, "step": 1479 }, { "epoch": 0.18827121231395497, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.9354329109191895, "learning_rate": 6.269605765154726e-07, "loss": 0.47, "mean_token_accuracy": 0.8461174368858337, "num_tokens": 56497448.0, "step": 1480 }, { "epoch": 0.18839842259254547, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.950448036193848, "learning_rate": 6.273844849512505e-07, "loss": 0.4724, "mean_token_accuracy": 0.8453062772750854, "num_tokens": 56537731.0, "step": 1481 }, { "epoch": 0.188525632871136, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.907097339630127, "learning_rate": 6.278083933870284e-07, "loss": 0.4116, "mean_token_accuracy": 0.866686999797821, "num_tokens": 56580944.0, "step": 1482 }, { "epoch": 0.1886528431497265, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.9824652671813965, "learning_rate": 6.282323018228063e-07, "loss": 0.4439, "mean_token_accuracy": 0.8572258949279785, "num_tokens": 56616306.0, "step": 1483 }, { "epoch": 0.188780053428317, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.963212966918945, "learning_rate": 6.286562102585841e-07, "loss": 0.5233, "mean_token_accuracy": 0.8352808952331543, "num_tokens": 56661874.0, "step": 1484 }, { "epoch": 0.18890726370690752, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.970612525939941, "learning_rate": 6.29080118694362e-07, "loss": 0.419, "mean_token_accuracy": 0.8622207045555115, "num_tokens": 56696697.0, "step": 1485 }, { "epoch": 0.18903447398549802, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.9882941246032715, "learning_rate": 6.295040271301398e-07, "loss": 0.418, "mean_token_accuracy": 0.8619118928909302, "num_tokens": 56732581.0, "step": 1486 }, { "epoch": 0.18916168426408853, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.880880355834961, "learning_rate": 6.299279355659178e-07, "loss": 0.4671, "mean_token_accuracy": 0.8494062423706055, "num_tokens": 56777905.0, "step": 1487 }, { "epoch": 0.18928889454267905, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.996882915496826, "learning_rate": 6.303518440016956e-07, "loss": 0.4866, "mean_token_accuracy": 0.842009425163269, "num_tokens": 56812915.0, "step": 1488 }, { "epoch": 0.18941610482126955, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 7.008016109466553, "learning_rate": 6.307757524374735e-07, "loss": 0.4814, "mean_token_accuracy": 0.8471279144287109, "num_tokens": 56846907.0, "step": 1489 }, { "epoch": 0.18954331509986008, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 7.004766464233398, "learning_rate": 6.311996608732514e-07, "loss": 0.51, "mean_token_accuracy": 0.8404279351234436, "num_tokens": 56884614.0, "step": 1490 }, { "epoch": 0.18967052537845058, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 7.007361888885498, "learning_rate": 6.316235693090292e-07, "loss": 0.501, "mean_token_accuracy": 0.8363662958145142, "num_tokens": 56920508.0, "step": 1491 }, { "epoch": 0.18979773565704108, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 6.957495212554932, "learning_rate": 6.320474777448071e-07, "loss": 0.5044, "mean_token_accuracy": 0.8363876938819885, "num_tokens": 56957629.0, "step": 1492 }, { "epoch": 0.1899249459356316, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.995955467224121, "learning_rate": 6.324713861805849e-07, "loss": 0.4622, "mean_token_accuracy": 0.847686767578125, "num_tokens": 56997265.0, "step": 1493 }, { "epoch": 0.1900521562142221, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 7.024839401245117, "learning_rate": 6.328952946163628e-07, "loss": 0.5646, "mean_token_accuracy": 0.8233035802841187, "num_tokens": 57031519.0, "step": 1494 }, { "epoch": 0.1901793664928126, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 7.015881538391113, "learning_rate": 6.333192030521407e-07, "loss": 0.4354, "mean_token_accuracy": 0.8539788126945496, "num_tokens": 57067264.0, "step": 1495 }, { "epoch": 0.19030657677140314, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 6.97601842880249, "learning_rate": 6.337431114879186e-07, "loss": 0.4898, "mean_token_accuracy": 0.8428338766098022, "num_tokens": 57104906.0, "step": 1496 }, { "epoch": 0.19043378704999364, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.997134685516357, "learning_rate": 6.341670199236965e-07, "loss": 0.4569, "mean_token_accuracy": 0.8521068692207336, "num_tokens": 57142534.0, "step": 1497 }, { "epoch": 0.19056099732858414, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 6.929884910583496, "learning_rate": 6.345909283594744e-07, "loss": 0.4382, "mean_token_accuracy": 0.8570278882980347, "num_tokens": 57188735.0, "step": 1498 }, { "epoch": 0.19068820760717467, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 7.006924152374268, "learning_rate": 6.350148367952522e-07, "loss": 0.4898, "mean_token_accuracy": 0.8356547355651855, "num_tokens": 57224610.0, "step": 1499 }, { "epoch": 0.19081541788576517, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 7.037030220031738, "learning_rate": 6.354387452310301e-07, "loss": 0.4449, "mean_token_accuracy": 0.8517290353775024, "num_tokens": 57259553.0, "step": 1500 }, { "epoch": 0.19094262816435567, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.967024803161621, "learning_rate": 6.358626536668079e-07, "loss": 0.449, "mean_token_accuracy": 0.8545898199081421, "num_tokens": 57295819.0, "step": 1501 }, { "epoch": 0.1910698384429462, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.938467502593994, "learning_rate": 6.362865621025858e-07, "loss": 0.4376, "mean_token_accuracy": 0.8587350845336914, "num_tokens": 57339110.0, "step": 1502 }, { "epoch": 0.1911970487215367, "ewc_loss": 0.006988525390625, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 7.02093505859375, "learning_rate": 6.367104705383637e-07, "loss": 0.4663, "mean_token_accuracy": 0.8483456969261169, "num_tokens": 57377311.0, "step": 1503 }, { "epoch": 0.1913242590001272, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 6.980368137359619, "learning_rate": 6.371343789741416e-07, "loss": 0.4783, "mean_token_accuracy": 0.8411051630973816, "num_tokens": 57419281.0, "step": 1504 }, { "epoch": 0.19145146927871773, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 7.005598068237305, "learning_rate": 6.375582874099195e-07, "loss": 0.4663, "mean_token_accuracy": 0.8485015630722046, "num_tokens": 57458134.0, "step": 1505 }, { "epoch": 0.19157867955730823, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 7.056726455688477, "learning_rate": 6.379821958456974e-07, "loss": 0.5204, "mean_token_accuracy": 0.8326974511146545, "num_tokens": 57492931.0, "step": 1506 }, { "epoch": 0.19170588983589873, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 7.028772354125977, "learning_rate": 6.384061042814751e-07, "loss": 0.4909, "mean_token_accuracy": 0.840005099773407, "num_tokens": 57529581.0, "step": 1507 }, { "epoch": 0.19183310011448926, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.000466823577881, "learning_rate": 6.38830012717253e-07, "loss": 0.4794, "mean_token_accuracy": 0.8450338840484619, "num_tokens": 57570302.0, "step": 1508 }, { "epoch": 0.19196031039307976, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 6.972651481628418, "learning_rate": 6.392539211530309e-07, "loss": 0.4902, "mean_token_accuracy": 0.8392568826675415, "num_tokens": 57615974.0, "step": 1509 }, { "epoch": 0.19208752067167026, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.0196099281311035, "learning_rate": 6.396778295888087e-07, "loss": 0.4552, "mean_token_accuracy": 0.8543808460235596, "num_tokens": 57653737.0, "step": 1510 }, { "epoch": 0.1922147309502608, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 7.0019354820251465, "learning_rate": 6.401017380245867e-07, "loss": 0.4883, "mean_token_accuracy": 0.8404693603515625, "num_tokens": 57693431.0, "step": 1511 }, { "epoch": 0.1923419412288513, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 6.978147506713867, "learning_rate": 6.405256464603645e-07, "loss": 0.4731, "mean_token_accuracy": 0.8455318212509155, "num_tokens": 57736413.0, "step": 1512 }, { "epoch": 0.1924691515074418, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 7.038843154907227, "learning_rate": 6.409495548961425e-07, "loss": 0.4576, "mean_token_accuracy": 0.8528984785079956, "num_tokens": 57773293.0, "step": 1513 }, { "epoch": 0.19259636178603232, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.058994293212891, "learning_rate": 6.413734633319203e-07, "loss": 0.4698, "mean_token_accuracy": 0.848433256149292, "num_tokens": 57809294.0, "step": 1514 }, { "epoch": 0.19272357206462282, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 7.070450305938721, "learning_rate": 6.417973717676981e-07, "loss": 0.4414, "mean_token_accuracy": 0.8534577488899231, "num_tokens": 57843610.0, "step": 1515 }, { "epoch": 0.19285078234321335, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 6.980645656585693, "learning_rate": 6.42221280203476e-07, "loss": 0.502, "mean_token_accuracy": 0.8349482417106628, "num_tokens": 57893098.0, "step": 1516 }, { "epoch": 0.19297799262180385, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.089888572692871, "learning_rate": 6.426451886392539e-07, "loss": 0.528, "mean_token_accuracy": 0.8345742225646973, "num_tokens": 57933098.0, "step": 1517 }, { "epoch": 0.19310520290039435, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 7.020162105560303, "learning_rate": 6.430690970750317e-07, "loss": 0.4594, "mean_token_accuracy": 0.8506630659103394, "num_tokens": 57971502.0, "step": 1518 }, { "epoch": 0.19323241317898487, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.0980048179626465, "learning_rate": 6.434930055108097e-07, "loss": 0.5091, "mean_token_accuracy": 0.8346933722496033, "num_tokens": 58001557.0, "step": 1519 }, { "epoch": 0.19335962345757537, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 7.081509590148926, "learning_rate": 6.439169139465875e-07, "loss": 0.5178, "mean_token_accuracy": 0.8322726488113403, "num_tokens": 58038058.0, "step": 1520 }, { "epoch": 0.19348683373616588, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.000877380371094, "learning_rate": 6.443408223823655e-07, "loss": 0.4859, "mean_token_accuracy": 0.8414942026138306, "num_tokens": 58073138.0, "step": 1521 }, { "epoch": 0.1936140440147564, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.044101715087891, "learning_rate": 6.447647308181432e-07, "loss": 0.4776, "mean_token_accuracy": 0.8443300724029541, "num_tokens": 58111793.0, "step": 1522 }, { "epoch": 0.1937412542933469, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.0029449462890625, "learning_rate": 6.451886392539211e-07, "loss": 0.4654, "mean_token_accuracy": 0.8475346565246582, "num_tokens": 58152911.0, "step": 1523 }, { "epoch": 0.1938684645719374, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 7.0561370849609375, "learning_rate": 6.45612547689699e-07, "loss": 0.4832, "mean_token_accuracy": 0.8446786999702454, "num_tokens": 58192095.0, "step": 1524 }, { "epoch": 0.19399567485052793, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 7.056245803833008, "learning_rate": 6.460364561254769e-07, "loss": 0.438, "mean_token_accuracy": 0.8556843400001526, "num_tokens": 58232651.0, "step": 1525 }, { "epoch": 0.19412288512911843, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 6.993262767791748, "learning_rate": 6.464603645612547e-07, "loss": 0.4361, "mean_token_accuracy": 0.8580572605133057, "num_tokens": 58270802.0, "step": 1526 }, { "epoch": 0.19425009540770893, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.069833755493164, "learning_rate": 6.468842729970327e-07, "loss": 0.468, "mean_token_accuracy": 0.8502954244613647, "num_tokens": 58313179.0, "step": 1527 }, { "epoch": 0.19437730568629946, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.071418762207031, "learning_rate": 6.473081814328105e-07, "loss": 0.5076, "mean_token_accuracy": 0.8405126333236694, "num_tokens": 58354728.0, "step": 1528 }, { "epoch": 0.19450451596488996, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.080159664154053, "learning_rate": 6.477320898685885e-07, "loss": 0.4909, "mean_token_accuracy": 0.8437392711639404, "num_tokens": 58398649.0, "step": 1529 }, { "epoch": 0.19463172624348046, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.069064140319824, "learning_rate": 6.481559983043662e-07, "loss": 0.5021, "mean_token_accuracy": 0.8380130529403687, "num_tokens": 58438261.0, "step": 1530 }, { "epoch": 0.194758936522071, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 7.033871173858643, "learning_rate": 6.48579906740144e-07, "loss": 0.4469, "mean_token_accuracy": 0.8533447980880737, "num_tokens": 58479141.0, "step": 1531 }, { "epoch": 0.1948861468006615, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.074349403381348, "learning_rate": 6.49003815175922e-07, "loss": 0.4151, "mean_token_accuracy": 0.8645418882369995, "num_tokens": 58515700.0, "step": 1532 }, { "epoch": 0.195013357079252, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.078317165374756, "learning_rate": 6.494277236116998e-07, "loss": 0.4219, "mean_token_accuracy": 0.8622051477432251, "num_tokens": 58549825.0, "step": 1533 }, { "epoch": 0.19514056735784252, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.058265686035156, "learning_rate": 6.498516320474777e-07, "loss": 0.5619, "mean_token_accuracy": 0.8236963748931885, "num_tokens": 58596110.0, "step": 1534 }, { "epoch": 0.19526777763643302, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 7.084643840789795, "learning_rate": 6.502755404832556e-07, "loss": 0.4721, "mean_token_accuracy": 0.8447879552841187, "num_tokens": 58632410.0, "step": 1535 }, { "epoch": 0.19539498791502352, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.090407848358154, "learning_rate": 6.506994489190335e-07, "loss": 0.5337, "mean_token_accuracy": 0.8332692980766296, "num_tokens": 58671339.0, "step": 1536 }, { "epoch": 0.19552219819361405, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 7.049661636352539, "learning_rate": 6.511233573548114e-07, "loss": 0.4337, "mean_token_accuracy": 0.8570626378059387, "num_tokens": 58711786.0, "step": 1537 }, { "epoch": 0.19564940847220455, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.119966983795166, "learning_rate": 6.515472657905892e-07, "loss": 0.4774, "mean_token_accuracy": 0.8453686237335205, "num_tokens": 58743600.0, "step": 1538 }, { "epoch": 0.19577661875079505, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.114398002624512, "learning_rate": 6.51971174226367e-07, "loss": 0.5333, "mean_token_accuracy": 0.8288543820381165, "num_tokens": 58780622.0, "step": 1539 }, { "epoch": 0.19590382902938558, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.096482276916504, "learning_rate": 6.52395082662145e-07, "loss": 0.5262, "mean_token_accuracy": 0.830832839012146, "num_tokens": 58819370.0, "step": 1540 }, { "epoch": 0.19603103930797608, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.092981338500977, "learning_rate": 6.528189910979228e-07, "loss": 0.4507, "mean_token_accuracy": 0.8533412218093872, "num_tokens": 58856270.0, "step": 1541 }, { "epoch": 0.1961582495865666, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.034832954406738, "learning_rate": 6.532428995337007e-07, "loss": 0.4681, "mean_token_accuracy": 0.8474615812301636, "num_tokens": 58898702.0, "step": 1542 }, { "epoch": 0.1962854598651571, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 7.112652778625488, "learning_rate": 6.536668079694786e-07, "loss": 0.4893, "mean_token_accuracy": 0.8390172123908997, "num_tokens": 58934602.0, "step": 1543 }, { "epoch": 0.1964126701437476, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.137750625610352, "learning_rate": 6.540907164052565e-07, "loss": 0.476, "mean_token_accuracy": 0.8447638750076294, "num_tokens": 58977766.0, "step": 1544 }, { "epoch": 0.19653988042233814, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 7.032696723937988, "learning_rate": 6.545146248410343e-07, "loss": 0.4854, "mean_token_accuracy": 0.8427994251251221, "num_tokens": 59019649.0, "step": 1545 }, { "epoch": 0.19666709070092864, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.1209235191345215, "learning_rate": 6.549385332768122e-07, "loss": 0.4975, "mean_token_accuracy": 0.8409289121627808, "num_tokens": 59060543.0, "step": 1546 }, { "epoch": 0.19679430097951914, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.149038791656494, "learning_rate": 6.5536244171259e-07, "loss": 0.4965, "mean_token_accuracy": 0.8383437395095825, "num_tokens": 59100049.0, "step": 1547 }, { "epoch": 0.19692151125810967, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.13894510269165, "learning_rate": 6.55786350148368e-07, "loss": 0.5027, "mean_token_accuracy": 0.8372846245765686, "num_tokens": 59137370.0, "step": 1548 }, { "epoch": 0.19704872153670017, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.1105546951293945, "learning_rate": 6.562102585841458e-07, "loss": 0.4553, "mean_token_accuracy": 0.8472787141799927, "num_tokens": 59176196.0, "step": 1549 }, { "epoch": 0.19717593181529067, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.095983028411865, "learning_rate": 6.566341670199236e-07, "loss": 0.4537, "mean_token_accuracy": 0.8535475730895996, "num_tokens": 59216215.0, "step": 1550 }, { "epoch": 0.1973031420938812, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.141280651092529, "learning_rate": 6.570580754557016e-07, "loss": 0.4702, "mean_token_accuracy": 0.8469645977020264, "num_tokens": 59254173.0, "step": 1551 }, { "epoch": 0.1974303523724717, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.190234184265137, "learning_rate": 6.574819838914794e-07, "loss": 0.4423, "mean_token_accuracy": 0.8556926250457764, "num_tokens": 59288496.0, "step": 1552 }, { "epoch": 0.1975575626510622, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.131724834442139, "learning_rate": 6.579058923272573e-07, "loss": 0.5006, "mean_token_accuracy": 0.8369936943054199, "num_tokens": 59329043.0, "step": 1553 }, { "epoch": 0.19768477292965272, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 7.131283760070801, "learning_rate": 6.583298007630351e-07, "loss": 0.4913, "mean_token_accuracy": 0.84133380651474, "num_tokens": 59360743.0, "step": 1554 }, { "epoch": 0.19781198320824323, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.112732410430908, "learning_rate": 6.58753709198813e-07, "loss": 0.488, "mean_token_accuracy": 0.8412891626358032, "num_tokens": 59404167.0, "step": 1555 }, { "epoch": 0.19793919348683373, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 7.0821685791015625, "learning_rate": 6.591776176345909e-07, "loss": 0.4558, "mean_token_accuracy": 0.8489351272583008, "num_tokens": 59447770.0, "step": 1556 }, { "epoch": 0.19806640376542425, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.1165361404418945, "learning_rate": 6.596015260703688e-07, "loss": 0.465, "mean_token_accuracy": 0.852507472038269, "num_tokens": 59494989.0, "step": 1557 }, { "epoch": 0.19819361404401475, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.121615886688232, "learning_rate": 6.600254345061466e-07, "loss": 0.4525, "mean_token_accuracy": 0.8544068336486816, "num_tokens": 59534926.0, "step": 1558 }, { "epoch": 0.19832082432260525, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.178651332855225, "learning_rate": 6.604493429419246e-07, "loss": 0.5076, "mean_token_accuracy": 0.8368619680404663, "num_tokens": 59577832.0, "step": 1559 }, { "epoch": 0.19844803460119578, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.154130458831787, "learning_rate": 6.608732513777023e-07, "loss": 0.5152, "mean_token_accuracy": 0.84160977602005, "num_tokens": 59624995.0, "step": 1560 }, { "epoch": 0.19857524487978628, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.126739501953125, "learning_rate": 6.612971598134803e-07, "loss": 0.4681, "mean_token_accuracy": 0.8473641872406006, "num_tokens": 59663000.0, "step": 1561 }, { "epoch": 0.19870245515837678, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.186947822570801, "learning_rate": 6.617210682492581e-07, "loss": 0.5222, "mean_token_accuracy": 0.83234703540802, "num_tokens": 59700241.0, "step": 1562 }, { "epoch": 0.1988296654369673, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.134184837341309, "learning_rate": 6.62144976685036e-07, "loss": 0.4414, "mean_token_accuracy": 0.8575090169906616, "num_tokens": 59737244.0, "step": 1563 }, { "epoch": 0.1989568757155578, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.147545337677002, "learning_rate": 6.625688851208139e-07, "loss": 0.491, "mean_token_accuracy": 0.841833770275116, "num_tokens": 59775538.0, "step": 1564 }, { "epoch": 0.19908408599414834, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.1825642585754395, "learning_rate": 6.629927935565918e-07, "loss": 0.4858, "mean_token_accuracy": 0.839972734451294, "num_tokens": 59810373.0, "step": 1565 }, { "epoch": 0.19921129627273884, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.179150104522705, "learning_rate": 6.634167019923696e-07, "loss": 0.5175, "mean_token_accuracy": 0.8371108174324036, "num_tokens": 59849992.0, "step": 1566 }, { "epoch": 0.19933850655132934, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.171532154083252, "learning_rate": 6.638406104281476e-07, "loss": 0.4785, "mean_token_accuracy": 0.844296395778656, "num_tokens": 59882552.0, "step": 1567 }, { "epoch": 0.19946571682991987, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.436471939086914e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.121704578399658, "learning_rate": 6.642645188639253e-07, "loss": 0.4538, "mean_token_accuracy": 0.8509465456008911, "num_tokens": 59925090.0, "step": 1568 }, { "epoch": 0.19959292710851037, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 1.436471939086914e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.188612461090088, "learning_rate": 6.646884272997032e-07, "loss": 0.5017, "mean_token_accuracy": 0.8383256196975708, "num_tokens": 59965406.0, "step": 1569 }, { "epoch": 0.19972013738710087, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.436471939086914e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.190911769866943, "learning_rate": 6.651123357354811e-07, "loss": 0.4883, "mean_token_accuracy": 0.8422900438308716, "num_tokens": 60001770.0, "step": 1570 }, { "epoch": 0.1998473476656914, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.436471939086914e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.187215328216553, "learning_rate": 6.655362441712589e-07, "loss": 0.5049, "mean_token_accuracy": 0.835182785987854, "num_tokens": 60036911.0, "step": 1571 }, { "epoch": 0.1999745579442819, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 1.436471939086914e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.1592020988464355, "learning_rate": 6.659601526070369e-07, "loss": 0.5328, "mean_token_accuracy": 0.8319979906082153, "num_tokens": 60080210.0, "step": 1572 }, { "epoch": 0.2001017682228724, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.221800804138184, "learning_rate": 6.663840610428147e-07, "loss": 0.5669, "mean_token_accuracy": 0.8240227699279785, "num_tokens": 60111533.0, "step": 1573 }, { "epoch": 0.20022897850146293, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.16945219039917, "learning_rate": 6.668079694785926e-07, "loss": 0.444, "mean_token_accuracy": 0.8558105230331421, "num_tokens": 60150733.0, "step": 1574 }, { "epoch": 0.20035618878005343, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.197063446044922, "learning_rate": 6.672318779143704e-07, "loss": 0.4638, "mean_token_accuracy": 0.8503453135490417, "num_tokens": 60185034.0, "step": 1575 }, { "epoch": 0.20048339905864393, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.215431213378906, "learning_rate": 6.676557863501483e-07, "loss": 0.4946, "mean_token_accuracy": 0.8401718735694885, "num_tokens": 60216865.0, "step": 1576 }, { "epoch": 0.20061060933723446, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.152549743652344, "learning_rate": 6.680796947859262e-07, "loss": 0.4829, "mean_token_accuracy": 0.8434690833091736, "num_tokens": 60257685.0, "step": 1577 }, { "epoch": 0.20073781961582496, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.224222660064697, "learning_rate": 6.685036032217041e-07, "loss": 0.478, "mean_token_accuracy": 0.8454234004020691, "num_tokens": 60294258.0, "step": 1578 }, { "epoch": 0.20086502989441546, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.194915294647217, "learning_rate": 6.689275116574819e-07, "loss": 0.4677, "mean_token_accuracy": 0.850911557674408, "num_tokens": 60330319.0, "step": 1579 }, { "epoch": 0.200992240173006, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.184264183044434, "learning_rate": 6.693514200932599e-07, "loss": 0.4846, "mean_token_accuracy": 0.8459487557411194, "num_tokens": 60370482.0, "step": 1580 }, { "epoch": 0.2011194504515965, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.174969673156738, "learning_rate": 6.697753285290377e-07, "loss": 0.4894, "mean_token_accuracy": 0.8412500023841858, "num_tokens": 60412436.0, "step": 1581 }, { "epoch": 0.201246660730187, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.212629795074463, "learning_rate": 6.701992369648156e-07, "loss": 0.4136, "mean_token_accuracy": 0.8648535013198853, "num_tokens": 60448982.0, "step": 1582 }, { "epoch": 0.20137387100877752, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.180870056152344, "learning_rate": 6.706231454005934e-07, "loss": 0.4306, "mean_token_accuracy": 0.8575818538665771, "num_tokens": 60489878.0, "step": 1583 }, { "epoch": 0.20150108128736802, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.241854667663574, "learning_rate": 6.710470538363713e-07, "loss": 0.4665, "mean_token_accuracy": 0.8509969711303711, "num_tokens": 60528979.0, "step": 1584 }, { "epoch": 0.20162829156595852, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.26861047744751, "learning_rate": 6.714709622721492e-07, "loss": 0.4945, "mean_token_accuracy": 0.8395954370498657, "num_tokens": 60561670.0, "step": 1585 }, { "epoch": 0.20175550184454905, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.197049140930176, "learning_rate": 6.718948707079271e-07, "loss": 0.4687, "mean_token_accuracy": 0.8460038900375366, "num_tokens": 60602061.0, "step": 1586 }, { "epoch": 0.20188271212313955, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.199367523193359, "learning_rate": 6.723187791437049e-07, "loss": 0.4718, "mean_token_accuracy": 0.8468013405799866, "num_tokens": 60639805.0, "step": 1587 }, { "epoch": 0.20200992240173005, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.217149257659912, "learning_rate": 6.727426875794829e-07, "loss": 0.4987, "mean_token_accuracy": 0.839671790599823, "num_tokens": 60681816.0, "step": 1588 }, { "epoch": 0.20213713268032057, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.232202053070068, "learning_rate": 6.731665960152607e-07, "loss": 0.4827, "mean_token_accuracy": 0.8412953615188599, "num_tokens": 60726343.0, "step": 1589 }, { "epoch": 0.20226434295891108, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.214820861816406, "learning_rate": 6.735905044510385e-07, "loss": 0.5272, "mean_token_accuracy": 0.8243417739868164, "num_tokens": 60764814.0, "step": 1590 }, { "epoch": 0.2023915532375016, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.2622270584106445, "learning_rate": 6.740144128868164e-07, "loss": 0.4941, "mean_token_accuracy": 0.8408955931663513, "num_tokens": 60802384.0, "step": 1591 }, { "epoch": 0.2025187635160921, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.24230432510376, "learning_rate": 6.744383213225942e-07, "loss": 0.5204, "mean_token_accuracy": 0.831179678440094, "num_tokens": 60842085.0, "step": 1592 }, { "epoch": 0.2026459737946826, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.195311546325684, "learning_rate": 6.748622297583722e-07, "loss": 0.4561, "mean_token_accuracy": 0.8476892113685608, "num_tokens": 60874835.0, "step": 1593 }, { "epoch": 0.20277318407327313, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.246241569519043, "learning_rate": 6.7528613819415e-07, "loss": 0.4821, "mean_token_accuracy": 0.841379702091217, "num_tokens": 60910682.0, "step": 1594 }, { "epoch": 0.20290039435186363, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.212458610534668, "learning_rate": 6.757100466299279e-07, "loss": 0.4791, "mean_token_accuracy": 0.8474252820014954, "num_tokens": 60951352.0, "step": 1595 }, { "epoch": 0.20302760463045413, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.3415021896362305, "learning_rate": 6.761339550657058e-07, "loss": 0.5067, "mean_token_accuracy": 0.8337419629096985, "num_tokens": 60979055.0, "step": 1596 }, { "epoch": 0.20315481490904466, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.257289409637451, "learning_rate": 6.765578635014837e-07, "loss": 0.4856, "mean_token_accuracy": 0.8451158404350281, "num_tokens": 61017437.0, "step": 1597 }, { "epoch": 0.20328202518763516, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.248746871948242, "learning_rate": 6.769817719372614e-07, "loss": 0.4613, "mean_token_accuracy": 0.849960207939148, "num_tokens": 61052286.0, "step": 1598 }, { "epoch": 0.20340923546622566, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.270480155944824, "learning_rate": 6.774056803730394e-07, "loss": 0.5066, "mean_token_accuracy": 0.8357254266738892, "num_tokens": 61086110.0, "step": 1599 }, { "epoch": 0.2035364457448162, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.259214878082275, "learning_rate": 6.778295888088172e-07, "loss": 0.4783, "mean_token_accuracy": 0.8448303937911987, "num_tokens": 61124181.0, "step": 1600 }, { "epoch": 0.2036636560234067, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.279449462890625, "learning_rate": 6.782534972445952e-07, "loss": 0.4983, "mean_token_accuracy": 0.8369994163513184, "num_tokens": 61160881.0, "step": 1601 }, { "epoch": 0.2037908663019972, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.276287078857422, "learning_rate": 6.78677405680373e-07, "loss": 0.4965, "mean_token_accuracy": 0.8425275683403015, "num_tokens": 61200755.0, "step": 1602 }, { "epoch": 0.20391807658058772, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.246854782104492, "learning_rate": 6.791013141161509e-07, "loss": 0.4867, "mean_token_accuracy": 0.8402560949325562, "num_tokens": 61240045.0, "step": 1603 }, { "epoch": 0.20404528685917822, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.21954870223999, "learning_rate": 6.795252225519288e-07, "loss": 0.5119, "mean_token_accuracy": 0.8328850269317627, "num_tokens": 61282832.0, "step": 1604 }, { "epoch": 0.20417249713776872, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.2380194664001465, "learning_rate": 6.799491309877067e-07, "loss": 0.4859, "mean_token_accuracy": 0.8401907086372375, "num_tokens": 61319841.0, "step": 1605 }, { "epoch": 0.20429970741635925, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.301189422607422, "learning_rate": 6.803730394234844e-07, "loss": 0.4544, "mean_token_accuracy": 0.8528978824615479, "num_tokens": 61352732.0, "step": 1606 }, { "epoch": 0.20442691769494975, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.281613349914551, "learning_rate": 6.807969478592624e-07, "loss": 0.4878, "mean_token_accuracy": 0.8442221879959106, "num_tokens": 61388324.0, "step": 1607 }, { "epoch": 0.20455412797354025, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.227711200714111, "learning_rate": 6.812208562950402e-07, "loss": 0.5115, "mean_token_accuracy": 0.8405650854110718, "num_tokens": 61431222.0, "step": 1608 }, { "epoch": 0.20468133825213078, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.214051723480225, "learning_rate": 6.816447647308182e-07, "loss": 0.5412, "mean_token_accuracy": 0.8264721632003784, "num_tokens": 61475617.0, "step": 1609 }, { "epoch": 0.20480854853072128, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.259093761444092, "learning_rate": 6.82068673166596e-07, "loss": 0.4682, "mean_token_accuracy": 0.8494743704795837, "num_tokens": 61511338.0, "step": 1610 }, { "epoch": 0.20493575880931178, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.310496807098389, "learning_rate": 6.824925816023738e-07, "loss": 0.4897, "mean_token_accuracy": 0.8397798538208008, "num_tokens": 61549093.0, "step": 1611 }, { "epoch": 0.2050629690879023, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.193272113800049, "learning_rate": 6.829164900381518e-07, "loss": 0.458, "mean_token_accuracy": 0.8513062596321106, "num_tokens": 61593843.0, "step": 1612 }, { "epoch": 0.2051901793664928, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.336028099060059, "learning_rate": 6.833403984739295e-07, "loss": 0.4692, "mean_token_accuracy": 0.8461147546768188, "num_tokens": 61629362.0, "step": 1613 }, { "epoch": 0.2053173896450833, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.329262733459473, "learning_rate": 6.837643069097074e-07, "loss": 0.55, "mean_token_accuracy": 0.8244306445121765, "num_tokens": 61667037.0, "step": 1614 }, { "epoch": 0.20544459992367384, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.305150032043457, "learning_rate": 6.841882153454853e-07, "loss": 0.4671, "mean_token_accuracy": 0.8459391593933105, "num_tokens": 61701976.0, "step": 1615 }, { "epoch": 0.20557181020226434, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.278329849243164, "learning_rate": 6.846121237812632e-07, "loss": 0.436, "mean_token_accuracy": 0.8572550415992737, "num_tokens": 61739068.0, "step": 1616 }, { "epoch": 0.20569902048085487, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.259873390197754, "learning_rate": 6.850360322170411e-07, "loss": 0.4522, "mean_token_accuracy": 0.8531985282897949, "num_tokens": 61781515.0, "step": 1617 }, { "epoch": 0.20582623075944537, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.346436023712158, "learning_rate": 6.85459940652819e-07, "loss": 0.4657, "mean_token_accuracy": 0.8460782766342163, "num_tokens": 61815201.0, "step": 1618 }, { "epoch": 0.20595344103803587, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.308553218841553, "learning_rate": 6.858838490885968e-07, "loss": 0.5225, "mean_token_accuracy": 0.8309204578399658, "num_tokens": 61856639.0, "step": 1619 }, { "epoch": 0.2060806513166264, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.335418701171875, "learning_rate": 6.863077575243748e-07, "loss": 0.5347, "mean_token_accuracy": 0.8309539556503296, "num_tokens": 61894313.0, "step": 1620 }, { "epoch": 0.2062078615952169, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.279575824737549, "learning_rate": 6.867316659601525e-07, "loss": 0.4704, "mean_token_accuracy": 0.8452907800674438, "num_tokens": 61930748.0, "step": 1621 }, { "epoch": 0.2063350718738074, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.293771266937256, "learning_rate": 6.871555743959304e-07, "loss": 0.4856, "mean_token_accuracy": 0.8445662260055542, "num_tokens": 61972002.0, "step": 1622 }, { "epoch": 0.20646228215239792, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.236481189727783, "learning_rate": 6.875794828317083e-07, "loss": 0.4612, "mean_token_accuracy": 0.851027250289917, "num_tokens": 62010141.0, "step": 1623 }, { "epoch": 0.20658949243098843, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.335978984832764, "learning_rate": 6.880033912674862e-07, "loss": 0.4497, "mean_token_accuracy": 0.855441153049469, "num_tokens": 62055907.0, "step": 1624 }, { "epoch": 0.20671670270957893, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.340686321258545, "learning_rate": 6.884272997032641e-07, "loss": 0.5563, "mean_token_accuracy": 0.8217974901199341, "num_tokens": 62095885.0, "step": 1625 }, { "epoch": 0.20684391298816945, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.291059494018555, "learning_rate": 6.88851208139042e-07, "loss": 0.4487, "mean_token_accuracy": 0.8559207320213318, "num_tokens": 62130497.0, "step": 1626 }, { "epoch": 0.20697112326675995, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.288655757904053, "learning_rate": 6.892751165748198e-07, "loss": 0.5082, "mean_token_accuracy": 0.8344072103500366, "num_tokens": 62170296.0, "step": 1627 }, { "epoch": 0.20709833354535045, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.299846172332764, "learning_rate": 6.896990250105978e-07, "loss": 0.4459, "mean_token_accuracy": 0.8586412668228149, "num_tokens": 62210168.0, "step": 1628 }, { "epoch": 0.20722554382394098, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.335046291351318, "learning_rate": 6.901229334463755e-07, "loss": 0.4708, "mean_token_accuracy": 0.8457887172698975, "num_tokens": 62245251.0, "step": 1629 }, { "epoch": 0.20735275410253148, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.285698413848877, "learning_rate": 6.905468418821534e-07, "loss": 0.4697, "mean_token_accuracy": 0.8520087003707886, "num_tokens": 62286124.0, "step": 1630 }, { "epoch": 0.20747996438112198, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.314981460571289, "learning_rate": 6.909707503179313e-07, "loss": 0.474, "mean_token_accuracy": 0.8433812856674194, "num_tokens": 62325604.0, "step": 1631 }, { "epoch": 0.2076071746597125, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.354041576385498, "learning_rate": 6.913946587537091e-07, "loss": 0.4699, "mean_token_accuracy": 0.8469959497451782, "num_tokens": 62364467.0, "step": 1632 }, { "epoch": 0.207734384938303, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.323786735534668, "learning_rate": 6.918185671894871e-07, "loss": 0.5435, "mean_token_accuracy": 0.8258674144744873, "num_tokens": 62402075.0, "step": 1633 }, { "epoch": 0.2078615952168935, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.326213359832764, "learning_rate": 6.922424756252649e-07, "loss": 0.5108, "mean_token_accuracy": 0.842597246170044, "num_tokens": 62437802.0, "step": 1634 }, { "epoch": 0.20798880549548404, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.305487155914307, "learning_rate": 6.926663840610428e-07, "loss": 0.4293, "mean_token_accuracy": 0.8589011430740356, "num_tokens": 62476176.0, "step": 1635 }, { "epoch": 0.20811601577407454, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.3292951583862305, "learning_rate": 6.930902924968206e-07, "loss": 0.4741, "mean_token_accuracy": 0.8427956104278564, "num_tokens": 62514378.0, "step": 1636 }, { "epoch": 0.20824322605266504, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.277967929840088, "learning_rate": 6.935142009325985e-07, "loss": 0.4958, "mean_token_accuracy": 0.8398256301879883, "num_tokens": 62555302.0, "step": 1637 }, { "epoch": 0.20837043633125557, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.291682243347168, "learning_rate": 6.939381093683764e-07, "loss": 0.5144, "mean_token_accuracy": 0.8338297605514526, "num_tokens": 62594356.0, "step": 1638 }, { "epoch": 0.20849764660984607, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.331343173980713, "learning_rate": 6.943620178041543e-07, "loss": 0.4323, "mean_token_accuracy": 0.8604455590248108, "num_tokens": 62632903.0, "step": 1639 }, { "epoch": 0.2086248568884366, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.354000568389893, "learning_rate": 6.947859262399321e-07, "loss": 0.4826, "mean_token_accuracy": 0.8501484990119934, "num_tokens": 62668979.0, "step": 1640 }, { "epoch": 0.2087520671670271, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.426429748535156, "learning_rate": 6.952098346757101e-07, "loss": 0.4684, "mean_token_accuracy": 0.8471159934997559, "num_tokens": 62705867.0, "step": 1641 }, { "epoch": 0.2088792774456176, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.321774005889893, "learning_rate": 6.956337431114879e-07, "loss": 0.4642, "mean_token_accuracy": 0.8509321212768555, "num_tokens": 62744559.0, "step": 1642 }, { "epoch": 0.20900648772420813, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.316792964935303, "learning_rate": 6.960576515472658e-07, "loss": 0.4949, "mean_token_accuracy": 0.8407410383224487, "num_tokens": 62785221.0, "step": 1643 }, { "epoch": 0.20913369800279863, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.438033103942871, "learning_rate": 6.964815599830436e-07, "loss": 0.4847, "mean_token_accuracy": 0.8434735536575317, "num_tokens": 62820095.0, "step": 1644 }, { "epoch": 0.20926090828138913, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.355241775512695, "learning_rate": 6.969054684188215e-07, "loss": 0.5354, "mean_token_accuracy": 0.8248887062072754, "num_tokens": 62863414.0, "step": 1645 }, { "epoch": 0.20938811855997966, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.295522689819336, "learning_rate": 6.973293768545994e-07, "loss": 0.4457, "mean_token_accuracy": 0.8548001050949097, "num_tokens": 62904647.0, "step": 1646 }, { "epoch": 0.20951532883857016, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.309711456298828, "learning_rate": 6.977532852903773e-07, "loss": 0.4788, "mean_token_accuracy": 0.8433443307876587, "num_tokens": 62949476.0, "step": 1647 }, { "epoch": 0.20964253911716066, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.44524621963501, "learning_rate": 6.981771937261551e-07, "loss": 0.5314, "mean_token_accuracy": 0.8308046460151672, "num_tokens": 62984784.0, "step": 1648 }, { "epoch": 0.2097697493957512, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.339729309082031, "learning_rate": 6.986011021619331e-07, "loss": 0.4859, "mean_token_accuracy": 0.8434237241744995, "num_tokens": 63024395.0, "step": 1649 }, { "epoch": 0.2098969596743417, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.3646159172058105, "learning_rate": 6.990250105977109e-07, "loss": 0.5019, "mean_token_accuracy": 0.839655339717865, "num_tokens": 63061886.0, "step": 1650 }, { "epoch": 0.2100241699529322, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.353572845458984, "learning_rate": 6.994489190334886e-07, "loss": 0.4696, "mean_token_accuracy": 0.8461179137229919, "num_tokens": 63104150.0, "step": 1651 }, { "epoch": 0.21015138023152272, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.44426155090332, "learning_rate": 6.998728274692666e-07, "loss": 0.4336, "mean_token_accuracy": 0.8586854934692383, "num_tokens": 63144346.0, "step": 1652 }, { "epoch": 0.21027859051011322, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.393630504608154, "learning_rate": 7.002967359050444e-07, "loss": 0.4759, "mean_token_accuracy": 0.8520091772079468, "num_tokens": 63180042.0, "step": 1653 }, { "epoch": 0.21040580078870372, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.340670108795166, "learning_rate": 7.007206443408224e-07, "loss": 0.4256, "mean_token_accuracy": 0.85981285572052, "num_tokens": 63216570.0, "step": 1654 }, { "epoch": 0.21053301106729425, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.365527629852295, "learning_rate": 7.011445527766002e-07, "loss": 0.4774, "mean_token_accuracy": 0.8449685573577881, "num_tokens": 63253457.0, "step": 1655 }, { "epoch": 0.21066022134588475, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.439180374145508, "learning_rate": 7.015684612123781e-07, "loss": 0.4625, "mean_token_accuracy": 0.8484142422676086, "num_tokens": 63295430.0, "step": 1656 }, { "epoch": 0.21078743162447525, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.36733341217041, "learning_rate": 7.01992369648156e-07, "loss": 0.4844, "mean_token_accuracy": 0.8454890251159668, "num_tokens": 63340118.0, "step": 1657 }, { "epoch": 0.21091464190306577, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.447895050048828, "learning_rate": 7.024162780839339e-07, "loss": 0.5154, "mean_token_accuracy": 0.8269170522689819, "num_tokens": 63372037.0, "step": 1658 }, { "epoch": 0.21104185218165628, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.3969011306762695, "learning_rate": 7.028401865197116e-07, "loss": 0.5059, "mean_token_accuracy": 0.8353917598724365, "num_tokens": 63415508.0, "step": 1659 }, { "epoch": 0.21116906246024678, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.361380577087402, "learning_rate": 7.032640949554896e-07, "loss": 0.4153, "mean_token_accuracy": 0.8607192039489746, "num_tokens": 63459320.0, "step": 1660 }, { "epoch": 0.2112962727388373, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.45527458190918, "learning_rate": 7.036880033912674e-07, "loss": 0.5147, "mean_token_accuracy": 0.8356570601463318, "num_tokens": 63499624.0, "step": 1661 }, { "epoch": 0.2114234830174278, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.3956098556518555, "learning_rate": 7.041119118270454e-07, "loss": 0.4711, "mean_token_accuracy": 0.8467165231704712, "num_tokens": 63536507.0, "step": 1662 }, { "epoch": 0.2115506932960183, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.396352767944336, "learning_rate": 7.045358202628232e-07, "loss": 0.5122, "mean_token_accuracy": 0.8363940119743347, "num_tokens": 63576483.0, "step": 1663 }, { "epoch": 0.21167790357460883, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.383677959442139, "learning_rate": 7.049597286986011e-07, "loss": 0.5018, "mean_token_accuracy": 0.8363471031188965, "num_tokens": 63615568.0, "step": 1664 }, { "epoch": 0.21180511385319933, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.421250820159912, "learning_rate": 7.05383637134379e-07, "loss": 0.4622, "mean_token_accuracy": 0.8495134115219116, "num_tokens": 63650023.0, "step": 1665 }, { "epoch": 0.21193232413178986, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.358194828033447, "learning_rate": 7.058075455701568e-07, "loss": 0.4927, "mean_token_accuracy": 0.8367444276809692, "num_tokens": 63690229.0, "step": 1666 }, { "epoch": 0.21205953441038036, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.403642177581787, "learning_rate": 7.062314540059346e-07, "loss": 0.4371, "mean_token_accuracy": 0.860105037689209, "num_tokens": 63728169.0, "step": 1667 }, { "epoch": 0.21218674468897086, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.453415870666504, "learning_rate": 7.066553624417126e-07, "loss": 0.4622, "mean_token_accuracy": 0.8476821184158325, "num_tokens": 63762816.0, "step": 1668 }, { "epoch": 0.2123139549675614, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.391155242919922, "learning_rate": 7.070792708774904e-07, "loss": 0.4838, "mean_token_accuracy": 0.8437908887863159, "num_tokens": 63804288.0, "step": 1669 }, { "epoch": 0.2124411652461519, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.368382930755615, "learning_rate": 7.075031793132684e-07, "loss": 0.505, "mean_token_accuracy": 0.8413084745407104, "num_tokens": 63851365.0, "step": 1670 }, { "epoch": 0.2125683755247424, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.425230503082275, "learning_rate": 7.079270877490462e-07, "loss": 0.4664, "mean_token_accuracy": 0.849951982498169, "num_tokens": 63886187.0, "step": 1671 }, { "epoch": 0.21269558580333292, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.455856800079346, "learning_rate": 7.08350996184824e-07, "loss": 0.5397, "mean_token_accuracy": 0.8274281024932861, "num_tokens": 63927187.0, "step": 1672 }, { "epoch": 0.21282279608192342, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.41119384765625, "learning_rate": 7.08774904620602e-07, "loss": 0.4604, "mean_token_accuracy": 0.8535932302474976, "num_tokens": 63964088.0, "step": 1673 }, { "epoch": 0.21295000636051392, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.4408345222473145, "learning_rate": 7.091988130563797e-07, "loss": 0.5046, "mean_token_accuracy": 0.8370176553726196, "num_tokens": 64004931.0, "step": 1674 }, { "epoch": 0.21307721663910445, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.357995510101318, "learning_rate": 7.096227214921576e-07, "loss": 0.4984, "mean_token_accuracy": 0.8385411500930786, "num_tokens": 64043966.0, "step": 1675 }, { "epoch": 0.21320442691769495, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.499124526977539, "learning_rate": 7.100466299279355e-07, "loss": 0.4811, "mean_token_accuracy": 0.845497190952301, "num_tokens": 64078886.0, "step": 1676 }, { "epoch": 0.21333163719628545, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.471938133239746, "learning_rate": 7.104705383637134e-07, "loss": 0.5047, "mean_token_accuracy": 0.8420617580413818, "num_tokens": 64117291.0, "step": 1677 }, { "epoch": 0.21345884747487598, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.402373790740967, "learning_rate": 7.108944467994913e-07, "loss": 0.462, "mean_token_accuracy": 0.8488600254058838, "num_tokens": 64156369.0, "step": 1678 }, { "epoch": 0.21358605775346648, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.476906776428223, "learning_rate": 7.113183552352692e-07, "loss": 0.5349, "mean_token_accuracy": 0.8239305019378662, "num_tokens": 64193646.0, "step": 1679 }, { "epoch": 0.21371326803205698, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.453482151031494, "learning_rate": 7.11742263671047e-07, "loss": 0.4277, "mean_token_accuracy": 0.8590134978294373, "num_tokens": 64231933.0, "step": 1680 }, { "epoch": 0.2138404783106475, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.5038604736328125, "learning_rate": 7.12166172106825e-07, "loss": 0.5595, "mean_token_accuracy": 0.8182892799377441, "num_tokens": 64267418.0, "step": 1681 }, { "epoch": 0.213967688589238, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.481791973114014, "learning_rate": 7.125900805426027e-07, "loss": 0.4409, "mean_token_accuracy": 0.8552665710449219, "num_tokens": 64302440.0, "step": 1682 }, { "epoch": 0.2140948988678285, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.4351372718811035, "learning_rate": 7.130139889783806e-07, "loss": 0.4795, "mean_token_accuracy": 0.8450342416763306, "num_tokens": 64338240.0, "step": 1683 }, { "epoch": 0.21422210914641904, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.454623222351074, "learning_rate": 7.134378974141585e-07, "loss": 0.471, "mean_token_accuracy": 0.8460919260978699, "num_tokens": 64375734.0, "step": 1684 }, { "epoch": 0.21434931942500954, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.505185127258301, "learning_rate": 7.138618058499364e-07, "loss": 0.4651, "mean_token_accuracy": 0.8492711782455444, "num_tokens": 64411401.0, "step": 1685 }, { "epoch": 0.21447652970360004, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.4397993087768555, "learning_rate": 7.142857142857143e-07, "loss": 0.4399, "mean_token_accuracy": 0.8581377267837524, "num_tokens": 64454719.0, "step": 1686 }, { "epoch": 0.21460373998219057, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.542929649353027, "learning_rate": 7.147096227214922e-07, "loss": 0.4645, "mean_token_accuracy": 0.8500608801841736, "num_tokens": 64495736.0, "step": 1687 }, { "epoch": 0.21473095026078107, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.500467777252197, "learning_rate": 7.1513353115727e-07, "loss": 0.4448, "mean_token_accuracy": 0.853911280632019, "num_tokens": 64532843.0, "step": 1688 }, { "epoch": 0.21485816053937157, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.47663688659668, "learning_rate": 7.155574395930479e-07, "loss": 0.4907, "mean_token_accuracy": 0.8373030424118042, "num_tokens": 64568950.0, "step": 1689 }, { "epoch": 0.2149853708179621, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.524147987365723, "learning_rate": 7.159813480288257e-07, "loss": 0.468, "mean_token_accuracy": 0.8453184366226196, "num_tokens": 64606878.0, "step": 1690 }, { "epoch": 0.2151125810965526, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.527624130249023, "learning_rate": 7.164052564646035e-07, "loss": 0.4357, "mean_token_accuracy": 0.8570549488067627, "num_tokens": 64643012.0, "step": 1691 }, { "epoch": 0.21523979137514312, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.454534530639648, "learning_rate": 7.168291649003815e-07, "loss": 0.4563, "mean_token_accuracy": 0.8547016978263855, "num_tokens": 64684444.0, "step": 1692 }, { "epoch": 0.21536700165373363, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.536665916442871, "learning_rate": 7.172530733361593e-07, "loss": 0.5055, "mean_token_accuracy": 0.8384870886802673, "num_tokens": 64722068.0, "step": 1693 }, { "epoch": 0.21549421193232413, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.468898773193359, "learning_rate": 7.176769817719373e-07, "loss": 0.4178, "mean_token_accuracy": 0.868800163269043, "num_tokens": 64761829.0, "step": 1694 }, { "epoch": 0.21562142221091465, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.529791355133057, "learning_rate": 7.181008902077151e-07, "loss": 0.4646, "mean_token_accuracy": 0.8498250842094421, "num_tokens": 64799185.0, "step": 1695 }, { "epoch": 0.21574863248950515, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.510441780090332, "learning_rate": 7.18524798643493e-07, "loss": 0.4938, "mean_token_accuracy": 0.8419486880302429, "num_tokens": 64839306.0, "step": 1696 }, { "epoch": 0.21587584276809565, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.481719493865967, "learning_rate": 7.189487070792708e-07, "loss": 0.4691, "mean_token_accuracy": 0.8503656387329102, "num_tokens": 64876011.0, "step": 1697 }, { "epoch": 0.21600305304668618, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.495131015777588, "learning_rate": 7.193726155150487e-07, "loss": 0.4684, "mean_token_accuracy": 0.8516055941581726, "num_tokens": 64918510.0, "step": 1698 }, { "epoch": 0.21613026332527668, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.462533950805664, "learning_rate": 7.197965239508265e-07, "loss": 0.4747, "mean_token_accuracy": 0.8466810584068298, "num_tokens": 64957156.0, "step": 1699 }, { "epoch": 0.21625747360386718, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.568127632141113, "learning_rate": 7.202204323866045e-07, "loss": 0.4729, "mean_token_accuracy": 0.8485345840454102, "num_tokens": 64998761.0, "step": 1700 }, { "epoch": 0.2163846838824577, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.500359535217285, "learning_rate": 7.206443408223823e-07, "loss": 0.4761, "mean_token_accuracy": 0.8445239067077637, "num_tokens": 65038849.0, "step": 1701 }, { "epoch": 0.2165118941610482, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.563933849334717, "learning_rate": 7.210682492581603e-07, "loss": 0.4561, "mean_token_accuracy": 0.8488729000091553, "num_tokens": 65071130.0, "step": 1702 }, { "epoch": 0.2166391044396387, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.587619781494141, "learning_rate": 7.214921576939381e-07, "loss": 0.4975, "mean_token_accuracy": 0.8357086777687073, "num_tokens": 65111619.0, "step": 1703 }, { "epoch": 0.21676631471822924, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.512408256530762, "learning_rate": 7.219160661297159e-07, "loss": 0.445, "mean_token_accuracy": 0.8528205156326294, "num_tokens": 65147537.0, "step": 1704 }, { "epoch": 0.21689352499681974, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.535229206085205, "learning_rate": 7.223399745654938e-07, "loss": 0.482, "mean_token_accuracy": 0.8431428074836731, "num_tokens": 65187254.0, "step": 1705 }, { "epoch": 0.21702073527541024, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.556042194366455, "learning_rate": 7.227638830012717e-07, "loss": 0.4936, "mean_token_accuracy": 0.8367100954055786, "num_tokens": 65226130.0, "step": 1706 }, { "epoch": 0.21714794555400077, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.574735164642334, "learning_rate": 7.231877914370495e-07, "loss": 0.4564, "mean_token_accuracy": 0.853706419467926, "num_tokens": 65261282.0, "step": 1707 }, { "epoch": 0.21727515583259127, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.548209190368652, "learning_rate": 7.236116998728275e-07, "loss": 0.5077, "mean_token_accuracy": 0.8359529972076416, "num_tokens": 65294850.0, "step": 1708 }, { "epoch": 0.21740236611118177, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.4808220863342285, "learning_rate": 7.240356083086053e-07, "loss": 0.4007, "mean_token_accuracy": 0.8670544028282166, "num_tokens": 65332323.0, "step": 1709 }, { "epoch": 0.2175295763897723, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.545614719390869, "learning_rate": 7.244595167443833e-07, "loss": 0.4939, "mean_token_accuracy": 0.8415699005126953, "num_tokens": 65376033.0, "step": 1710 }, { "epoch": 0.2176567866683628, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.504800319671631, "learning_rate": 7.248834251801611e-07, "loss": 0.4592, "mean_token_accuracy": 0.8525699377059937, "num_tokens": 65417875.0, "step": 1711 }, { "epoch": 0.2177839969469533, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.524781227111816, "learning_rate": 7.253073336159388e-07, "loss": 0.436, "mean_token_accuracy": 0.8569439649581909, "num_tokens": 65456675.0, "step": 1712 }, { "epoch": 0.21791120722554383, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.6026105880737305, "learning_rate": 7.257312420517168e-07, "loss": 0.428, "mean_token_accuracy": 0.8598263263702393, "num_tokens": 65487699.0, "step": 1713 }, { "epoch": 0.21803841750413433, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.560884475708008, "learning_rate": 7.261551504874946e-07, "loss": 0.4551, "mean_token_accuracy": 0.8481510877609253, "num_tokens": 65526143.0, "step": 1714 }, { "epoch": 0.21816562778272486, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.5058393478393555, "learning_rate": 7.265790589232725e-07, "loss": 0.4526, "mean_token_accuracy": 0.853229284286499, "num_tokens": 65562969.0, "step": 1715 }, { "epoch": 0.21829283806131536, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.522083282470703, "learning_rate": 7.270029673590504e-07, "loss": 0.4854, "mean_token_accuracy": 0.8454879522323608, "num_tokens": 65605999.0, "step": 1716 }, { "epoch": 0.21842004833990586, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.6032280921936035, "learning_rate": 7.274268757948283e-07, "loss": 0.5297, "mean_token_accuracy": 0.8308941125869751, "num_tokens": 65644877.0, "step": 1717 }, { "epoch": 0.2185472586184964, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.588933944702148, "learning_rate": 7.278507842306062e-07, "loss": 0.4506, "mean_token_accuracy": 0.8540748357772827, "num_tokens": 65678823.0, "step": 1718 }, { "epoch": 0.2186744688970869, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.551790714263916, "learning_rate": 7.282746926663841e-07, "loss": 0.5303, "mean_token_accuracy": 0.8312399983406067, "num_tokens": 65717185.0, "step": 1719 }, { "epoch": 0.2188016791756774, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.612891674041748, "learning_rate": 7.286986011021618e-07, "loss": 0.4506, "mean_token_accuracy": 0.8538190722465515, "num_tokens": 65754422.0, "step": 1720 }, { "epoch": 0.21892888945426792, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.4999823570251465, "learning_rate": 7.291225095379398e-07, "loss": 0.4869, "mean_token_accuracy": 0.8442232608795166, "num_tokens": 65799390.0, "step": 1721 }, { "epoch": 0.21905609973285842, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.5918731689453125, "learning_rate": 7.295464179737176e-07, "loss": 0.4787, "mean_token_accuracy": 0.848108172416687, "num_tokens": 65837566.0, "step": 1722 }, { "epoch": 0.21918331001144892, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.534457683563232, "learning_rate": 7.299703264094955e-07, "loss": 0.4637, "mean_token_accuracy": 0.8491360545158386, "num_tokens": 65874537.0, "step": 1723 }, { "epoch": 0.21931052029003945, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.5651397705078125, "learning_rate": 7.303942348452734e-07, "loss": 0.4897, "mean_token_accuracy": 0.8412300944328308, "num_tokens": 65915616.0, "step": 1724 }, { "epoch": 0.21943773056862995, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.583093643188477, "learning_rate": 7.308181432810513e-07, "loss": 0.4393, "mean_token_accuracy": 0.8554175496101379, "num_tokens": 65954263.0, "step": 1725 }, { "epoch": 0.21956494084722045, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.552115440368652, "learning_rate": 7.312420517168292e-07, "loss": 0.4718, "mean_token_accuracy": 0.8448092937469482, "num_tokens": 65992880.0, "step": 1726 }, { "epoch": 0.21969215112581097, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.605350494384766, "learning_rate": 7.31665960152607e-07, "loss": 0.4905, "mean_token_accuracy": 0.8382899761199951, "num_tokens": 66033958.0, "step": 1727 }, { "epoch": 0.21981936140440148, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.567694664001465, "learning_rate": 7.320898685883848e-07, "loss": 0.4616, "mean_token_accuracy": 0.8484430909156799, "num_tokens": 66072448.0, "step": 1728 }, { "epoch": 0.21994657168299198, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.5903520584106445, "learning_rate": 7.325137770241628e-07, "loss": 0.5726, "mean_token_accuracy": 0.8186962008476257, "num_tokens": 66112648.0, "step": 1729 }, { "epoch": 0.2200737819615825, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.570357322692871, "learning_rate": 7.329376854599406e-07, "loss": 0.4834, "mean_token_accuracy": 0.8446570038795471, "num_tokens": 66154791.0, "step": 1730 }, { "epoch": 0.220200992240173, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.588252544403076, "learning_rate": 7.333615938957184e-07, "loss": 0.4709, "mean_token_accuracy": 0.846863865852356, "num_tokens": 66192450.0, "step": 1731 }, { "epoch": 0.2203282025187635, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 7.638290882110596, "learning_rate": 7.337855023314964e-07, "loss": 0.4667, "mean_token_accuracy": 0.8466615676879883, "num_tokens": 66231173.0, "step": 1732 }, { "epoch": 0.22045541279735403, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.614384651184082, "learning_rate": 7.342094107672742e-07, "loss": 0.5196, "mean_token_accuracy": 0.8317690491676331, "num_tokens": 66272956.0, "step": 1733 }, { "epoch": 0.22058262307594453, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.605718612670898, "learning_rate": 7.346333192030522e-07, "loss": 0.4738, "mean_token_accuracy": 0.8507223725318909, "num_tokens": 66309450.0, "step": 1734 }, { "epoch": 0.22070983335453503, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.642091751098633, "learning_rate": 7.350572276388299e-07, "loss": 0.5234, "mean_token_accuracy": 0.8299949169158936, "num_tokens": 66350947.0, "step": 1735 }, { "epoch": 0.22083704363312556, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.629433631896973, "learning_rate": 7.354811360746078e-07, "loss": 0.4811, "mean_token_accuracy": 0.8424890637397766, "num_tokens": 66388364.0, "step": 1736 }, { "epoch": 0.22096425391171606, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 7.613266944885254, "learning_rate": 7.359050445103857e-07, "loss": 0.496, "mean_token_accuracy": 0.8387315273284912, "num_tokens": 66425029.0, "step": 1737 }, { "epoch": 0.22109146419030656, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 7.553646564483643, "learning_rate": 7.363289529461636e-07, "loss": 0.4718, "mean_token_accuracy": 0.8467522859573364, "num_tokens": 66463924.0, "step": 1738 }, { "epoch": 0.2212186744688971, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.576467990875244, "learning_rate": 7.367528613819415e-07, "loss": 0.4264, "mean_token_accuracy": 0.8602023124694824, "num_tokens": 66504824.0, "step": 1739 }, { "epoch": 0.2213458847474876, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.617527008056641, "learning_rate": 7.371767698177194e-07, "loss": 0.446, "mean_token_accuracy": 0.854097843170166, "num_tokens": 66539308.0, "step": 1740 }, { "epoch": 0.22147309502607812, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.614716053009033, "learning_rate": 7.376006782534972e-07, "loss": 0.4417, "mean_token_accuracy": 0.8569509387016296, "num_tokens": 66572785.0, "step": 1741 }, { "epoch": 0.22160030530466862, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.611660480499268, "learning_rate": 7.380245866892751e-07, "loss": 0.4589, "mean_token_accuracy": 0.8504006266593933, "num_tokens": 66607890.0, "step": 1742 }, { "epoch": 0.22172751558325912, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.59376335144043, "learning_rate": 7.384484951250529e-07, "loss": 0.4616, "mean_token_accuracy": 0.8520102500915527, "num_tokens": 66650553.0, "step": 1743 }, { "epoch": 0.22185472586184965, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.630606651306152, "learning_rate": 7.388724035608308e-07, "loss": 0.4567, "mean_token_accuracy": 0.8522729873657227, "num_tokens": 66685330.0, "step": 1744 }, { "epoch": 0.22198193614044015, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.640106678009033, "learning_rate": 7.392963119966087e-07, "loss": 0.4398, "mean_token_accuracy": 0.8553104996681213, "num_tokens": 66716816.0, "step": 1745 }, { "epoch": 0.22210914641903065, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.62455415725708, "learning_rate": 7.397202204323866e-07, "loss": 0.4843, "mean_token_accuracy": 0.8441356420516968, "num_tokens": 66754030.0, "step": 1746 }, { "epoch": 0.22223635669762118, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.625838756561279, "learning_rate": 7.401441288681645e-07, "loss": 0.442, "mean_token_accuracy": 0.8560149073600769, "num_tokens": 66793650.0, "step": 1747 }, { "epoch": 0.22236356697621168, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.637725353240967, "learning_rate": 7.405680373039424e-07, "loss": 0.4258, "mean_token_accuracy": 0.8635392189025879, "num_tokens": 66833709.0, "step": 1748 }, { "epoch": 0.22249077725480218, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.6401567459106445, "learning_rate": 7.409919457397202e-07, "loss": 0.4938, "mean_token_accuracy": 0.8393406867980957, "num_tokens": 66866456.0, "step": 1749 }, { "epoch": 0.2226179875333927, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.642570495605469, "learning_rate": 7.414158541754981e-07, "loss": 0.5198, "mean_token_accuracy": 0.8312854766845703, "num_tokens": 66905875.0, "step": 1750 }, { "epoch": 0.2227451978119832, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.788492202758789, "learning_rate": 7.418397626112759e-07, "loss": 0.4747, "mean_token_accuracy": 0.8533713221549988, "num_tokens": 66936009.0, "step": 1751 }, { "epoch": 0.2228724080905737, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.63101863861084, "learning_rate": 7.422636710470537e-07, "loss": 0.4779, "mean_token_accuracy": 0.8455735445022583, "num_tokens": 66978111.0, "step": 1752 }, { "epoch": 0.22299961836916424, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.631279468536377, "learning_rate": 7.426875794828317e-07, "loss": 0.4907, "mean_token_accuracy": 0.844390869140625, "num_tokens": 67015938.0, "step": 1753 }, { "epoch": 0.22312682864775474, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.673672199249268, "learning_rate": 7.431114879186095e-07, "loss": 0.5123, "mean_token_accuracy": 0.8351626396179199, "num_tokens": 67055456.0, "step": 1754 }, { "epoch": 0.22325403892634524, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.693726062774658, "learning_rate": 7.435353963543875e-07, "loss": 0.4699, "mean_token_accuracy": 0.8474641442298889, "num_tokens": 67096298.0, "step": 1755 }, { "epoch": 0.22338124920493577, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.627048969268799, "learning_rate": 7.439593047901653e-07, "loss": 0.4259, "mean_token_accuracy": 0.8609434962272644, "num_tokens": 67135280.0, "step": 1756 }, { "epoch": 0.22350845948352627, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.65596342086792, "learning_rate": 7.443832132259431e-07, "loss": 0.49, "mean_token_accuracy": 0.8394230604171753, "num_tokens": 67171936.0, "step": 1757 }, { "epoch": 0.22363566976211677, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.661981105804443, "learning_rate": 7.44807121661721e-07, "loss": 0.4783, "mean_token_accuracy": 0.8438214063644409, "num_tokens": 67209854.0, "step": 1758 }, { "epoch": 0.2237628800407073, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.686946392059326, "learning_rate": 7.452310300974989e-07, "loss": 0.4332, "mean_token_accuracy": 0.8540518283843994, "num_tokens": 67244369.0, "step": 1759 }, { "epoch": 0.2238900903192978, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.641702651977539, "learning_rate": 7.456549385332767e-07, "loss": 0.5114, "mean_token_accuracy": 0.8361897468566895, "num_tokens": 67283994.0, "step": 1760 }, { "epoch": 0.2240173005978883, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.6727986335754395, "learning_rate": 7.460788469690547e-07, "loss": 0.3975, "mean_token_accuracy": 0.8669922351837158, "num_tokens": 67323230.0, "step": 1761 }, { "epoch": 0.22414451087647883, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.657248497009277, "learning_rate": 7.465027554048325e-07, "loss": 0.5496, "mean_token_accuracy": 0.8261123299598694, "num_tokens": 67366198.0, "step": 1762 }, { "epoch": 0.22427172115506933, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.697439670562744, "learning_rate": 7.469266638406105e-07, "loss": 0.5339, "mean_token_accuracy": 0.8309192061424255, "num_tokens": 67410685.0, "step": 1763 }, { "epoch": 0.22439893143365983, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.659907817840576, "learning_rate": 7.473505722763883e-07, "loss": 0.4563, "mean_token_accuracy": 0.8534194827079773, "num_tokens": 67445818.0, "step": 1764 }, { "epoch": 0.22452614171225035, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.656183242797852, "learning_rate": 7.477744807121661e-07, "loss": 0.4989, "mean_token_accuracy": 0.8386288285255432, "num_tokens": 67483873.0, "step": 1765 }, { "epoch": 0.22465335199084085, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.682110786437988, "learning_rate": 7.48198389147944e-07, "loss": 0.4767, "mean_token_accuracy": 0.8500120639801025, "num_tokens": 67519524.0, "step": 1766 }, { "epoch": 0.22478056226943138, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.617507457733154, "learning_rate": 7.486222975837219e-07, "loss": 0.4614, "mean_token_accuracy": 0.8483269810676575, "num_tokens": 67560920.0, "step": 1767 }, { "epoch": 0.22490777254802188, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.6772356033325195, "learning_rate": 7.490462060194997e-07, "loss": 0.4325, "mean_token_accuracy": 0.8558710813522339, "num_tokens": 67594948.0, "step": 1768 }, { "epoch": 0.22503498282661238, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.677559852600098, "learning_rate": 7.494701144552777e-07, "loss": 0.5196, "mean_token_accuracy": 0.8325444459915161, "num_tokens": 67628867.0, "step": 1769 }, { "epoch": 0.2251621931052029, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.662100315093994, "learning_rate": 7.498940228910555e-07, "loss": 0.4804, "mean_token_accuracy": 0.842768669128418, "num_tokens": 67669762.0, "step": 1770 }, { "epoch": 0.2252894033837934, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.7043304443359375, "learning_rate": 7.503179313268335e-07, "loss": 0.4677, "mean_token_accuracy": 0.846030592918396, "num_tokens": 67704750.0, "step": 1771 }, { "epoch": 0.2254166136623839, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.600780963897705, "learning_rate": 7.507418397626113e-07, "loss": 0.5049, "mean_token_accuracy": 0.8368498086929321, "num_tokens": 67745181.0, "step": 1772 }, { "epoch": 0.22554382394097444, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.64291524887085, "learning_rate": 7.51165748198389e-07, "loss": 0.492, "mean_token_accuracy": 0.839144229888916, "num_tokens": 67785929.0, "step": 1773 }, { "epoch": 0.22567103421956494, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.682911396026611, "learning_rate": 7.51589656634167e-07, "loss": 0.4478, "mean_token_accuracy": 0.8501620292663574, "num_tokens": 67821887.0, "step": 1774 }, { "epoch": 0.22579824449815544, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.686059474945068, "learning_rate": 7.520135650699448e-07, "loss": 0.4824, "mean_token_accuracy": 0.8435000777244568, "num_tokens": 67859608.0, "step": 1775 }, { "epoch": 0.22592545477674597, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.639350414276123, "learning_rate": 7.524374735057227e-07, "loss": 0.4912, "mean_token_accuracy": 0.8414425849914551, "num_tokens": 67895287.0, "step": 1776 }, { "epoch": 0.22605266505533647, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.730433464050293, "learning_rate": 7.528613819415006e-07, "loss": 0.4499, "mean_token_accuracy": 0.8576380014419556, "num_tokens": 67930358.0, "step": 1777 }, { "epoch": 0.22617987533392697, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.687840461730957, "learning_rate": 7.532852903772785e-07, "loss": 0.4907, "mean_token_accuracy": 0.840261697769165, "num_tokens": 67966112.0, "step": 1778 }, { "epoch": 0.2263070856125175, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.823362350463867, "learning_rate": 7.537091988130564e-07, "loss": 0.4981, "mean_token_accuracy": 0.8357813954353333, "num_tokens": 68001297.0, "step": 1779 }, { "epoch": 0.226434295891108, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.695755481719971, "learning_rate": 7.541331072488342e-07, "loss": 0.4795, "mean_token_accuracy": 0.8453919887542725, "num_tokens": 68037346.0, "step": 1780 }, { "epoch": 0.2265615061696985, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.677180767059326, "learning_rate": 7.54557015684612e-07, "loss": 0.4781, "mean_token_accuracy": 0.8439866304397583, "num_tokens": 68076566.0, "step": 1781 }, { "epoch": 0.22668871644828903, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.730515480041504, "learning_rate": 7.5498092412039e-07, "loss": 0.4568, "mean_token_accuracy": 0.8542966842651367, "num_tokens": 68116834.0, "step": 1782 }, { "epoch": 0.22681592672687953, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.690345764160156, "learning_rate": 7.554048325561678e-07, "loss": 0.4306, "mean_token_accuracy": 0.8604041934013367, "num_tokens": 68155573.0, "step": 1783 }, { "epoch": 0.22694313700547003, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.718386173248291, "learning_rate": 7.558287409919457e-07, "loss": 0.5237, "mean_token_accuracy": 0.834122359752655, "num_tokens": 68193378.0, "step": 1784 }, { "epoch": 0.22707034728406056, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.738397121429443, "learning_rate": 7.562526494277236e-07, "loss": 0.4967, "mean_token_accuracy": 0.8365756869316101, "num_tokens": 68230183.0, "step": 1785 }, { "epoch": 0.22719755756265106, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.690305709838867, "learning_rate": 7.566765578635015e-07, "loss": 0.4661, "mean_token_accuracy": 0.8498948216438293, "num_tokens": 68270526.0, "step": 1786 }, { "epoch": 0.22732476784124156, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.7415618896484375, "learning_rate": 7.571004662992794e-07, "loss": 0.4917, "mean_token_accuracy": 0.845245361328125, "num_tokens": 68308690.0, "step": 1787 }, { "epoch": 0.2274519781198321, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.705496311187744, "learning_rate": 7.575243747350572e-07, "loss": 0.5342, "mean_token_accuracy": 0.8307434916496277, "num_tokens": 68348660.0, "step": 1788 }, { "epoch": 0.2275791883984226, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.755399227142334, "learning_rate": 7.57948283170835e-07, "loss": 0.4829, "mean_token_accuracy": 0.8470364809036255, "num_tokens": 68389129.0, "step": 1789 }, { "epoch": 0.2277063986770131, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.751518249511719, "learning_rate": 7.58372191606613e-07, "loss": 0.4747, "mean_token_accuracy": 0.8479299545288086, "num_tokens": 68423234.0, "step": 1790 }, { "epoch": 0.22783360895560362, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.793533802032471, "learning_rate": 7.587961000423908e-07, "loss": 0.4818, "mean_token_accuracy": 0.8455840945243835, "num_tokens": 68460834.0, "step": 1791 }, { "epoch": 0.22796081923419412, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.75772762298584, "learning_rate": 7.592200084781686e-07, "loss": 0.526, "mean_token_accuracy": 0.8352451324462891, "num_tokens": 68495605.0, "step": 1792 }, { "epoch": 0.22808802951278465, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.808261871337891, "learning_rate": 7.596439169139466e-07, "loss": 0.4727, "mean_token_accuracy": 0.8472371101379395, "num_tokens": 68530838.0, "step": 1793 }, { "epoch": 0.22821523979137515, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.74119758605957, "learning_rate": 7.600678253497244e-07, "loss": 0.4664, "mean_token_accuracy": 0.8512650728225708, "num_tokens": 68571225.0, "step": 1794 }, { "epoch": 0.22834245006996565, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.749808311462402, "learning_rate": 7.604917337855023e-07, "loss": 0.4482, "mean_token_accuracy": 0.8564989566802979, "num_tokens": 68608284.0, "step": 1795 }, { "epoch": 0.22846966034855618, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.793145656585693, "learning_rate": 7.609156422212801e-07, "loss": 0.4299, "mean_token_accuracy": 0.8589406609535217, "num_tokens": 68645943.0, "step": 1796 }, { "epoch": 0.22859687062714668, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.753866672515869, "learning_rate": 7.61339550657058e-07, "loss": 0.4529, "mean_token_accuracy": 0.8515908718109131, "num_tokens": 68684245.0, "step": 1797 }, { "epoch": 0.22872408090573718, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.817000389099121, "learning_rate": 7.617634590928359e-07, "loss": 0.4665, "mean_token_accuracy": 0.8512797355651855, "num_tokens": 68717532.0, "step": 1798 }, { "epoch": 0.2288512911843277, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.756701946258545, "learning_rate": 7.621873675286138e-07, "loss": 0.4855, "mean_token_accuracy": 0.8451250791549683, "num_tokens": 68759515.0, "step": 1799 }, { "epoch": 0.2289785014629182, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.765329837799072, "learning_rate": 7.626112759643916e-07, "loss": 0.4918, "mean_token_accuracy": 0.8397687673568726, "num_tokens": 68796969.0, "step": 1800 }, { "epoch": 0.2291057117415087, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.79094934463501, "learning_rate": 7.630351844001696e-07, "loss": 0.4729, "mean_token_accuracy": 0.8458172678947449, "num_tokens": 68831138.0, "step": 1801 }, { "epoch": 0.22923292202009923, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.738955020904541, "learning_rate": 7.634590928359474e-07, "loss": 0.4627, "mean_token_accuracy": 0.8520744442939758, "num_tokens": 68868741.0, "step": 1802 }, { "epoch": 0.22936013229868973, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.7272725105285645, "learning_rate": 7.638830012717253e-07, "loss": 0.4839, "mean_token_accuracy": 0.8423184156417847, "num_tokens": 68909191.0, "step": 1803 }, { "epoch": 0.22948734257728023, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.75864839553833, "learning_rate": 7.643069097075031e-07, "loss": 0.459, "mean_token_accuracy": 0.8493872284889221, "num_tokens": 68949261.0, "step": 1804 }, { "epoch": 0.22961455285587076, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.79594087600708, "learning_rate": 7.64730818143281e-07, "loss": 0.4333, "mean_token_accuracy": 0.8553817272186279, "num_tokens": 68990434.0, "step": 1805 }, { "epoch": 0.22974176313446126, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.711803436279297, "learning_rate": 7.651547265790589e-07, "loss": 0.5155, "mean_token_accuracy": 0.8347896337509155, "num_tokens": 69029482.0, "step": 1806 }, { "epoch": 0.22986897341305176, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.777618885040283, "learning_rate": 7.655786350148368e-07, "loss": 0.4956, "mean_token_accuracy": 0.8387775421142578, "num_tokens": 69069766.0, "step": 1807 }, { "epoch": 0.2299961836916423, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.790836334228516, "learning_rate": 7.660025434506146e-07, "loss": 0.4822, "mean_token_accuracy": 0.8443673253059387, "num_tokens": 69114592.0, "step": 1808 }, { "epoch": 0.2301233939702328, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.779151916503906, "learning_rate": 7.664264518863926e-07, "loss": 0.4711, "mean_token_accuracy": 0.8473246097564697, "num_tokens": 69152599.0, "step": 1809 }, { "epoch": 0.2302506042488233, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.729736804962158, "learning_rate": 7.668503603221704e-07, "loss": 0.5234, "mean_token_accuracy": 0.8290482759475708, "num_tokens": 69193157.0, "step": 1810 }, { "epoch": 0.23037781452741382, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.7747979164123535, "learning_rate": 7.672742687579483e-07, "loss": 0.4351, "mean_token_accuracy": 0.8558611869812012, "num_tokens": 69230681.0, "step": 1811 }, { "epoch": 0.23050502480600432, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.797773361206055, "learning_rate": 7.676981771937261e-07, "loss": 0.4141, "mean_token_accuracy": 0.8600540161132812, "num_tokens": 69261292.0, "step": 1812 }, { "epoch": 0.23063223508459482, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.732175350189209, "learning_rate": 7.681220856295039e-07, "loss": 0.4445, "mean_token_accuracy": 0.8551509976387024, "num_tokens": 69301795.0, "step": 1813 }, { "epoch": 0.23075944536318535, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.857542037963867, "learning_rate": 7.685459940652819e-07, "loss": 0.468, "mean_token_accuracy": 0.8449251651763916, "num_tokens": 69337038.0, "step": 1814 }, { "epoch": 0.23088665564177585, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 7.710411548614502, "learning_rate": 7.689699025010597e-07, "loss": 0.4649, "mean_token_accuracy": 0.8474931716918945, "num_tokens": 69379871.0, "step": 1815 }, { "epoch": 0.23101386592036638, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.852425575256348, "learning_rate": 7.693938109368376e-07, "loss": 0.4862, "mean_token_accuracy": 0.840451717376709, "num_tokens": 69418021.0, "step": 1816 }, { "epoch": 0.23114107619895688, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.8318328857421875, "learning_rate": 7.698177193726155e-07, "loss": 0.4653, "mean_token_accuracy": 0.8471653461456299, "num_tokens": 69455574.0, "step": 1817 }, { "epoch": 0.23126828647754738, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.769185543060303, "learning_rate": 7.702416278083933e-07, "loss": 0.4891, "mean_token_accuracy": 0.8420007228851318, "num_tokens": 69489822.0, "step": 1818 }, { "epoch": 0.2313954967561379, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.7955498695373535, "learning_rate": 7.706655362441712e-07, "loss": 0.4599, "mean_token_accuracy": 0.84873366355896, "num_tokens": 69532821.0, "step": 1819 }, { "epoch": 0.2315227070347284, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 7.8421735763549805, "learning_rate": 7.710894446799491e-07, "loss": 0.4439, "mean_token_accuracy": 0.8524335622787476, "num_tokens": 69565674.0, "step": 1820 }, { "epoch": 0.2316499173133189, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.739307880401611, "learning_rate": 7.715133531157269e-07, "loss": 0.4438, "mean_token_accuracy": 0.8577338457107544, "num_tokens": 69606118.0, "step": 1821 }, { "epoch": 0.23177712759190944, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.77836275100708, "learning_rate": 7.719372615515049e-07, "loss": 0.5139, "mean_token_accuracy": 0.8350135087966919, "num_tokens": 69650102.0, "step": 1822 }, { "epoch": 0.23190433787049994, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.767651557922363, "learning_rate": 7.723611699872827e-07, "loss": 0.4593, "mean_token_accuracy": 0.8487898707389832, "num_tokens": 69693440.0, "step": 1823 }, { "epoch": 0.23203154814909044, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.8486409187316895, "learning_rate": 7.727850784230606e-07, "loss": 0.4785, "mean_token_accuracy": 0.8403551578521729, "num_tokens": 69732651.0, "step": 1824 }, { "epoch": 0.23215875842768097, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.80462646484375, "learning_rate": 7.732089868588385e-07, "loss": 0.4937, "mean_token_accuracy": 0.8431049585342407, "num_tokens": 69769915.0, "step": 1825 }, { "epoch": 0.23228596870627147, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.77199125289917, "learning_rate": 7.736328952946163e-07, "loss": 0.4158, "mean_token_accuracy": 0.8651076555252075, "num_tokens": 69813943.0, "step": 1826 }, { "epoch": 0.23241317898486197, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.852704048156738, "learning_rate": 7.740568037303942e-07, "loss": 0.4458, "mean_token_accuracy": 0.85213303565979, "num_tokens": 69847039.0, "step": 1827 }, { "epoch": 0.2325403892634525, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.8018999099731445, "learning_rate": 7.744807121661721e-07, "loss": 0.4106, "mean_token_accuracy": 0.8677147626876831, "num_tokens": 69884588.0, "step": 1828 }, { "epoch": 0.232667599542043, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.831578731536865, "learning_rate": 7.749046206019499e-07, "loss": 0.4748, "mean_token_accuracy": 0.8459961414337158, "num_tokens": 69925552.0, "step": 1829 }, { "epoch": 0.2327948098206335, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.7789998054504395, "learning_rate": 7.753285290377279e-07, "loss": 0.4086, "mean_token_accuracy": 0.8648324012756348, "num_tokens": 69961776.0, "step": 1830 }, { "epoch": 0.23292202009922403, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.839802265167236, "learning_rate": 7.757524374735057e-07, "loss": 0.5152, "mean_token_accuracy": 0.8331066966056824, "num_tokens": 69999775.0, "step": 1831 }, { "epoch": 0.23304923037781453, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.7899346351623535, "learning_rate": 7.761763459092836e-07, "loss": 0.4255, "mean_token_accuracy": 0.8627442121505737, "num_tokens": 70037294.0, "step": 1832 }, { "epoch": 0.23317644065640503, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.83292818069458, "learning_rate": 7.766002543450614e-07, "loss": 0.4613, "mean_token_accuracy": 0.8508960008621216, "num_tokens": 70075014.0, "step": 1833 }, { "epoch": 0.23330365093499555, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.840670108795166, "learning_rate": 7.770241627808392e-07, "loss": 0.4185, "mean_token_accuracy": 0.8629677295684814, "num_tokens": 70112186.0, "step": 1834 }, { "epoch": 0.23343086121358606, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.82282829284668, "learning_rate": 7.774480712166172e-07, "loss": 0.4958, "mean_token_accuracy": 0.8398338556289673, "num_tokens": 70153285.0, "step": 1835 }, { "epoch": 0.23355807149217656, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.811681747436523, "learning_rate": 7.77871979652395e-07, "loss": 0.4782, "mean_token_accuracy": 0.8453020453453064, "num_tokens": 70193382.0, "step": 1836 }, { "epoch": 0.23368528177076708, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.874578475952148, "learning_rate": 7.782958880881729e-07, "loss": 0.4673, "mean_token_accuracy": 0.8486282825469971, "num_tokens": 70227279.0, "step": 1837 }, { "epoch": 0.23381249204935758, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.858309745788574, "learning_rate": 7.787197965239508e-07, "loss": 0.4304, "mean_token_accuracy": 0.8560808300971985, "num_tokens": 70264395.0, "step": 1838 }, { "epoch": 0.23393970232794808, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.943892955780029, "learning_rate": 7.791437049597287e-07, "loss": 0.5452, "mean_token_accuracy": 0.8252590894699097, "num_tokens": 70295016.0, "step": 1839 }, { "epoch": 0.2340669126065386, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.828818321228027, "learning_rate": 7.795676133955065e-07, "loss": 0.4888, "mean_token_accuracy": 0.8426089286804199, "num_tokens": 70332774.0, "step": 1840 }, { "epoch": 0.2341941228851291, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.878084659576416, "learning_rate": 7.799915218312844e-07, "loss": 0.4737, "mean_token_accuracy": 0.8460618257522583, "num_tokens": 70366880.0, "step": 1841 }, { "epoch": 0.23432133316371964, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.803947925567627, "learning_rate": 7.804154302670622e-07, "loss": 0.4496, "mean_token_accuracy": 0.8537291288375854, "num_tokens": 70405235.0, "step": 1842 }, { "epoch": 0.23444854344231014, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.808045864105225, "learning_rate": 7.808393387028402e-07, "loss": 0.4588, "mean_token_accuracy": 0.849469780921936, "num_tokens": 70446503.0, "step": 1843 }, { "epoch": 0.23457575372090064, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.884942054748535, "learning_rate": 7.81263247138618e-07, "loss": 0.5095, "mean_token_accuracy": 0.8348150849342346, "num_tokens": 70483514.0, "step": 1844 }, { "epoch": 0.23470296399949117, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.863236427307129, "learning_rate": 7.816871555743959e-07, "loss": 0.4712, "mean_token_accuracy": 0.8458433151245117, "num_tokens": 70520944.0, "step": 1845 }, { "epoch": 0.23483017427808167, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.8319172859191895, "learning_rate": 7.821110640101738e-07, "loss": 0.4944, "mean_token_accuracy": 0.8371176719665527, "num_tokens": 70564399.0, "step": 1846 }, { "epoch": 0.23495738455667217, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.867108345031738, "learning_rate": 7.825349724459517e-07, "loss": 0.4679, "mean_token_accuracy": 0.8488832712173462, "num_tokens": 70604000.0, "step": 1847 }, { "epoch": 0.2350845948352627, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.8729352951049805, "learning_rate": 7.829588808817294e-07, "loss": 0.4606, "mean_token_accuracy": 0.849801242351532, "num_tokens": 70645133.0, "step": 1848 }, { "epoch": 0.2352118051138532, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.868865013122559, "learning_rate": 7.833827893175074e-07, "loss": 0.4944, "mean_token_accuracy": 0.8336881995201111, "num_tokens": 70675824.0, "step": 1849 }, { "epoch": 0.2353390153924437, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.851699352264404, "learning_rate": 7.838066977532852e-07, "loss": 0.4575, "mean_token_accuracy": 0.8505874872207642, "num_tokens": 70714103.0, "step": 1850 }, { "epoch": 0.23546622567103423, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.825282573699951, "learning_rate": 7.842306061890632e-07, "loss": 0.5062, "mean_token_accuracy": 0.8377671241760254, "num_tokens": 70754265.0, "step": 1851 }, { "epoch": 0.23559343594962473, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.9093451499938965, "learning_rate": 7.84654514624841e-07, "loss": 0.4748, "mean_token_accuracy": 0.8435260057449341, "num_tokens": 70791368.0, "step": 1852 }, { "epoch": 0.23572064622821523, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.892426013946533, "learning_rate": 7.850784230606188e-07, "loss": 0.5194, "mean_token_accuracy": 0.8357036113739014, "num_tokens": 70827397.0, "step": 1853 }, { "epoch": 0.23584785650680576, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.895864963531494, "learning_rate": 7.855023314963968e-07, "loss": 0.4898, "mean_token_accuracy": 0.8430906534194946, "num_tokens": 70872503.0, "step": 1854 }, { "epoch": 0.23597506678539626, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.899625778198242, "learning_rate": 7.859262399321746e-07, "loss": 0.4893, "mean_token_accuracy": 0.8401683568954468, "num_tokens": 70914376.0, "step": 1855 }, { "epoch": 0.23610227706398676, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.883432388305664, "learning_rate": 7.863501483679524e-07, "loss": 0.4467, "mean_token_accuracy": 0.8544954657554626, "num_tokens": 70954004.0, "step": 1856 }, { "epoch": 0.2362294873425773, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.850790023803711, "learning_rate": 7.867740568037303e-07, "loss": 0.4468, "mean_token_accuracy": 0.851036012172699, "num_tokens": 70991765.0, "step": 1857 }, { "epoch": 0.2363566976211678, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.931162357330322, "learning_rate": 7.871979652395082e-07, "loss": 0.4756, "mean_token_accuracy": 0.8429686427116394, "num_tokens": 71025478.0, "step": 1858 }, { "epoch": 0.2364839078997583, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.8623738288879395, "learning_rate": 7.876218736752861e-07, "loss": 0.4307, "mean_token_accuracy": 0.8616714477539062, "num_tokens": 71067287.0, "step": 1859 }, { "epoch": 0.23661111817834882, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.8698835372924805, "learning_rate": 7.88045782111064e-07, "loss": 0.4998, "mean_token_accuracy": 0.8393368721008301, "num_tokens": 71108095.0, "step": 1860 }, { "epoch": 0.23673832845693932, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.84186315536499, "learning_rate": 7.884696905468418e-07, "loss": 0.434, "mean_token_accuracy": 0.8590956926345825, "num_tokens": 71150297.0, "step": 1861 }, { "epoch": 0.23686553873552982, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.889081954956055, "learning_rate": 7.888935989826198e-07, "loss": 0.4818, "mean_token_accuracy": 0.8407067060470581, "num_tokens": 71189205.0, "step": 1862 }, { "epoch": 0.23699274901412035, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.923999309539795, "learning_rate": 7.893175074183976e-07, "loss": 0.471, "mean_token_accuracy": 0.8462605476379395, "num_tokens": 71228141.0, "step": 1863 }, { "epoch": 0.23711995929271085, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.898698806762695, "learning_rate": 7.897414158541754e-07, "loss": 0.471, "mean_token_accuracy": 0.8471234440803528, "num_tokens": 71267854.0, "step": 1864 }, { "epoch": 0.23724716957130135, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.916056156158447, "learning_rate": 7.901653242899533e-07, "loss": 0.5064, "mean_token_accuracy": 0.8405442833900452, "num_tokens": 71305030.0, "step": 1865 }, { "epoch": 0.23737437984989188, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.923943996429443, "learning_rate": 7.905892327257312e-07, "loss": 0.462, "mean_token_accuracy": 0.8489906787872314, "num_tokens": 71342202.0, "step": 1866 }, { "epoch": 0.23750159012848238, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.883478164672852, "learning_rate": 7.910131411615091e-07, "loss": 0.4627, "mean_token_accuracy": 0.8515488505363464, "num_tokens": 71379702.0, "step": 1867 }, { "epoch": 0.2376288004070729, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.903422832489014, "learning_rate": 7.91437049597287e-07, "loss": 0.4641, "mean_token_accuracy": 0.8487033843994141, "num_tokens": 71417785.0, "step": 1868 }, { "epoch": 0.2377560106856634, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.902736186981201, "learning_rate": 7.918609580330648e-07, "loss": 0.5077, "mean_token_accuracy": 0.8360166549682617, "num_tokens": 71461358.0, "step": 1869 }, { "epoch": 0.2378832209642539, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.935920715332031, "learning_rate": 7.922848664688428e-07, "loss": 0.505, "mean_token_accuracy": 0.8385363817214966, "num_tokens": 71500059.0, "step": 1870 }, { "epoch": 0.23801043124284443, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.940381050109863, "learning_rate": 7.927087749046205e-07, "loss": 0.416, "mean_token_accuracy": 0.8642528057098389, "num_tokens": 71537043.0, "step": 1871 }, { "epoch": 0.23813764152143493, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.89060640335083, "learning_rate": 7.931326833403983e-07, "loss": 0.4516, "mean_token_accuracy": 0.8511816263198853, "num_tokens": 71570671.0, "step": 1872 }, { "epoch": 0.23826485180002543, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.933182716369629, "learning_rate": 7.935565917761763e-07, "loss": 0.461, "mean_token_accuracy": 0.8512012958526611, "num_tokens": 71614813.0, "step": 1873 }, { "epoch": 0.23839206207861596, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.934042453765869, "learning_rate": 7.939805002119541e-07, "loss": 0.4768, "mean_token_accuracy": 0.8437816500663757, "num_tokens": 71650441.0, "step": 1874 }, { "epoch": 0.23851927235720646, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.9739089012146, "learning_rate": 7.944044086477321e-07, "loss": 0.4826, "mean_token_accuracy": 0.8445721864700317, "num_tokens": 71690425.0, "step": 1875 }, { "epoch": 0.23864648263579696, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.906493186950684, "learning_rate": 7.948283170835099e-07, "loss": 0.4303, "mean_token_accuracy": 0.8576214909553528, "num_tokens": 71727349.0, "step": 1876 }, { "epoch": 0.2387736929143875, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.923054218292236, "learning_rate": 7.952522255192878e-07, "loss": 0.3967, "mean_token_accuracy": 0.8698257207870483, "num_tokens": 71770386.0, "step": 1877 }, { "epoch": 0.238900903192978, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.950517654418945, "learning_rate": 7.956761339550657e-07, "loss": 0.4279, "mean_token_accuracy": 0.8596609830856323, "num_tokens": 71808946.0, "step": 1878 }, { "epoch": 0.2390281134715685, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 8.049318313598633, "learning_rate": 7.961000423908435e-07, "loss": 0.4695, "mean_token_accuracy": 0.8539725542068481, "num_tokens": 71848605.0, "step": 1879 }, { "epoch": 0.23915532375015902, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.987536430358887, "learning_rate": 7.965239508266214e-07, "loss": 0.4481, "mean_token_accuracy": 0.8532335758209229, "num_tokens": 71886771.0, "step": 1880 }, { "epoch": 0.23928253402874952, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.924742221832275, "learning_rate": 7.969478592623993e-07, "loss": 0.4256, "mean_token_accuracy": 0.8580551147460938, "num_tokens": 71924905.0, "step": 1881 }, { "epoch": 0.23940974430734002, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.955389022827148, "learning_rate": 7.973717676981771e-07, "loss": 0.4521, "mean_token_accuracy": 0.853569746017456, "num_tokens": 71964513.0, "step": 1882 }, { "epoch": 0.23953695458593055, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.945944786071777, "learning_rate": 7.977956761339551e-07, "loss": 0.4225, "mean_token_accuracy": 0.8609792590141296, "num_tokens": 72000918.0, "step": 1883 }, { "epoch": 0.23966416486452105, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.948256492614746, "learning_rate": 7.982195845697329e-07, "loss": 0.552, "mean_token_accuracy": 0.8224212527275085, "num_tokens": 72040840.0, "step": 1884 }, { "epoch": 0.23979137514311155, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.995703220367432, "learning_rate": 7.986434930055108e-07, "loss": 0.4254, "mean_token_accuracy": 0.858971357345581, "num_tokens": 72076757.0, "step": 1885 }, { "epoch": 0.23991858542170208, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.979479789733887, "learning_rate": 7.990674014412886e-07, "loss": 0.4818, "mean_token_accuracy": 0.845958948135376, "num_tokens": 72120760.0, "step": 1886 }, { "epoch": 0.24004579570029258, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.992249011993408, "learning_rate": 7.994913098770665e-07, "loss": 0.4685, "mean_token_accuracy": 0.8465149998664856, "num_tokens": 72159985.0, "step": 1887 }, { "epoch": 0.24017300597888308, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.923201084136963, "learning_rate": 7.999152183128444e-07, "loss": 0.502, "mean_token_accuracy": 0.8368550539016724, "num_tokens": 72196377.0, "step": 1888 }, { "epoch": 0.2403002162574736, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.962347030639648, "learning_rate": 8.003391267486223e-07, "loss": 0.4499, "mean_token_accuracy": 0.8535645008087158, "num_tokens": 72236585.0, "step": 1889 }, { "epoch": 0.2404274265360641, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 8.02607250213623, "learning_rate": 8.007630351844001e-07, "loss": 0.4624, "mean_token_accuracy": 0.8488503694534302, "num_tokens": 72268450.0, "step": 1890 }, { "epoch": 0.24055463681465464, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.980419158935547, "learning_rate": 8.011869436201781e-07, "loss": 0.4472, "mean_token_accuracy": 0.85382479429245, "num_tokens": 72304092.0, "step": 1891 }, { "epoch": 0.24068184709324514, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.95063591003418, "learning_rate": 8.016108520559559e-07, "loss": 0.4733, "mean_token_accuracy": 0.844897449016571, "num_tokens": 72347312.0, "step": 1892 }, { "epoch": 0.24080905737183564, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.953610420227051, "learning_rate": 8.020347604917338e-07, "loss": 0.4367, "mean_token_accuracy": 0.8572034239768982, "num_tokens": 72382135.0, "step": 1893 }, { "epoch": 0.24093626765042617, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.002340316772461, "learning_rate": 8.024586689275116e-07, "loss": 0.463, "mean_token_accuracy": 0.8490725755691528, "num_tokens": 72421526.0, "step": 1894 }, { "epoch": 0.24106347792901667, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.979818344116211, "learning_rate": 8.028825773632894e-07, "loss": 0.4649, "mean_token_accuracy": 0.850711464881897, "num_tokens": 72458592.0, "step": 1895 }, { "epoch": 0.24119068820760717, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.957425117492676, "learning_rate": 8.033064857990674e-07, "loss": 0.396, "mean_token_accuracy": 0.8683716058731079, "num_tokens": 72494734.0, "step": 1896 }, { "epoch": 0.2413178984861977, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.976442813873291, "learning_rate": 8.037303942348452e-07, "loss": 0.4457, "mean_token_accuracy": 0.854714035987854, "num_tokens": 72531560.0, "step": 1897 }, { "epoch": 0.2414451087647882, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.098167419433594, "learning_rate": 8.041543026706231e-07, "loss": 0.4438, "mean_token_accuracy": 0.8562332391738892, "num_tokens": 72570144.0, "step": 1898 }, { "epoch": 0.2415723190433787, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.017789840698242, "learning_rate": 8.04578211106401e-07, "loss": 0.4529, "mean_token_accuracy": 0.8498187065124512, "num_tokens": 72606738.0, "step": 1899 }, { "epoch": 0.24169952932196923, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.993666172027588, "learning_rate": 8.050021195421789e-07, "loss": 0.4968, "mean_token_accuracy": 0.8391115665435791, "num_tokens": 72646230.0, "step": 1900 }, { "epoch": 0.24182673960055973, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.01937484741211, "learning_rate": 8.054260279779567e-07, "loss": 0.4654, "mean_token_accuracy": 0.8458324670791626, "num_tokens": 72681811.0, "step": 1901 }, { "epoch": 0.24195394987915023, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.035886764526367, "learning_rate": 8.058499364137346e-07, "loss": 0.502, "mean_token_accuracy": 0.8373570442199707, "num_tokens": 72721805.0, "step": 1902 }, { "epoch": 0.24208116015774075, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.993110656738281, "learning_rate": 8.062738448495124e-07, "loss": 0.4155, "mean_token_accuracy": 0.862835168838501, "num_tokens": 72763142.0, "step": 1903 }, { "epoch": 0.24220837043633126, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.9773173332214355, "learning_rate": 8.066977532852904e-07, "loss": 0.4068, "mean_token_accuracy": 0.8706375360488892, "num_tokens": 72802047.0, "step": 1904 }, { "epoch": 0.24233558071492176, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.098292350769043, "learning_rate": 8.071216617210682e-07, "loss": 0.4844, "mean_token_accuracy": 0.8427289128303528, "num_tokens": 72834470.0, "step": 1905 }, { "epoch": 0.24246279099351228, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.033949851989746, "learning_rate": 8.075455701568461e-07, "loss": 0.4543, "mean_token_accuracy": 0.8521889448165894, "num_tokens": 72873101.0, "step": 1906 }, { "epoch": 0.24259000127210278, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.083969116210938, "learning_rate": 8.07969478592624e-07, "loss": 0.4347, "mean_token_accuracy": 0.8608947396278381, "num_tokens": 72915504.0, "step": 1907 }, { "epoch": 0.24271721155069328, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.036566734313965, "learning_rate": 8.083933870284019e-07, "loss": 0.409, "mean_token_accuracy": 0.8667204976081848, "num_tokens": 72947179.0, "step": 1908 }, { "epoch": 0.2428444218292838, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.08743667602539, "learning_rate": 8.088172954641796e-07, "loss": 0.449, "mean_token_accuracy": 0.8542491793632507, "num_tokens": 72983313.0, "step": 1909 }, { "epoch": 0.2429716321078743, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.105277061462402, "learning_rate": 8.092412038999576e-07, "loss": 0.4984, "mean_token_accuracy": 0.8398211002349854, "num_tokens": 73014218.0, "step": 1910 }, { "epoch": 0.24309884238646481, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.033414840698242, "learning_rate": 8.096651123357354e-07, "loss": 0.5077, "mean_token_accuracy": 0.8335403800010681, "num_tokens": 73048282.0, "step": 1911 }, { "epoch": 0.24322605266505534, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.049959182739258, "learning_rate": 8.100890207715134e-07, "loss": 0.4944, "mean_token_accuracy": 0.8407274484634399, "num_tokens": 73084390.0, "step": 1912 }, { "epoch": 0.24335326294364584, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.050568580627441, "learning_rate": 8.105129292072912e-07, "loss": 0.4617, "mean_token_accuracy": 0.8499047160148621, "num_tokens": 73123296.0, "step": 1913 }, { "epoch": 0.24348047322223634, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.071617126464844, "learning_rate": 8.10936837643069e-07, "loss": 0.5062, "mean_token_accuracy": 0.8382766246795654, "num_tokens": 73157533.0, "step": 1914 }, { "epoch": 0.24360768350082687, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.047274589538574, "learning_rate": 8.11360746078847e-07, "loss": 0.4409, "mean_token_accuracy": 0.8548858165740967, "num_tokens": 73195314.0, "step": 1915 }, { "epoch": 0.24373489377941737, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.026532173156738, "learning_rate": 8.117846545146248e-07, "loss": 0.4091, "mean_token_accuracy": 0.8636928796768188, "num_tokens": 73236763.0, "step": 1916 }, { "epoch": 0.2438621040580079, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.097174644470215, "learning_rate": 8.122085629504026e-07, "loss": 0.4972, "mean_token_accuracy": 0.8368962407112122, "num_tokens": 73276719.0, "step": 1917 }, { "epoch": 0.2439893143365984, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.037375450134277, "learning_rate": 8.126324713861805e-07, "loss": 0.4635, "mean_token_accuracy": 0.8514267206192017, "num_tokens": 73318594.0, "step": 1918 }, { "epoch": 0.2441165246151889, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.073062896728516, "learning_rate": 8.130563798219584e-07, "loss": 0.4387, "mean_token_accuracy": 0.858405351638794, "num_tokens": 73358760.0, "step": 1919 }, { "epoch": 0.24424373489377943, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.123512268066406, "learning_rate": 8.134802882577363e-07, "loss": 0.4816, "mean_token_accuracy": 0.847396731376648, "num_tokens": 73398623.0, "step": 1920 }, { "epoch": 0.24437094517236993, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.981194496154785, "learning_rate": 8.139041966935142e-07, "loss": 0.4636, "mean_token_accuracy": 0.8511689305305481, "num_tokens": 73446369.0, "step": 1921 }, { "epoch": 0.24449815545096043, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.157899856567383, "learning_rate": 8.14328105129292e-07, "loss": 0.5233, "mean_token_accuracy": 0.8361618518829346, "num_tokens": 73483152.0, "step": 1922 }, { "epoch": 0.24462536572955096, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.081195831298828, "learning_rate": 8.1475201356507e-07, "loss": 0.4777, "mean_token_accuracy": 0.8461340665817261, "num_tokens": 73519179.0, "step": 1923 }, { "epoch": 0.24475257600814146, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.07422161102295, "learning_rate": 8.151759220008477e-07, "loss": 0.4496, "mean_token_accuracy": 0.8517354726791382, "num_tokens": 73561947.0, "step": 1924 }, { "epoch": 0.24487978628673196, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.049721717834473, "learning_rate": 8.155998304366256e-07, "loss": 0.4646, "mean_token_accuracy": 0.8527002334594727, "num_tokens": 73595841.0, "step": 1925 }, { "epoch": 0.2450069965653225, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 8.158485412597656, "learning_rate": 8.160237388724035e-07, "loss": 0.4907, "mean_token_accuracy": 0.8384211659431458, "num_tokens": 73629716.0, "step": 1926 }, { "epoch": 0.245134206843913, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.112378120422363, "learning_rate": 8.164476473081814e-07, "loss": 0.4779, "mean_token_accuracy": 0.8453419208526611, "num_tokens": 73667285.0, "step": 1927 }, { "epoch": 0.2452614171225035, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.090981483459473, "learning_rate": 8.168715557439593e-07, "loss": 0.4223, "mean_token_accuracy": 0.862450122833252, "num_tokens": 73708059.0, "step": 1928 }, { "epoch": 0.24538862740109402, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.121560096740723, "learning_rate": 8.172954641797372e-07, "loss": 0.4352, "mean_token_accuracy": 0.8569892644882202, "num_tokens": 73748553.0, "step": 1929 }, { "epoch": 0.24551583767968452, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.05951976776123, "learning_rate": 8.17719372615515e-07, "loss": 0.5348, "mean_token_accuracy": 0.8277434706687927, "num_tokens": 73796295.0, "step": 1930 }, { "epoch": 0.24564304795827502, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.119949340820312, "learning_rate": 8.18143281051293e-07, "loss": 0.5141, "mean_token_accuracy": 0.8364812731742859, "num_tokens": 73835960.0, "step": 1931 }, { "epoch": 0.24577025823686555, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.10644817352295, "learning_rate": 8.185671894870707e-07, "loss": 0.5302, "mean_token_accuracy": 0.8295698761940002, "num_tokens": 73875302.0, "step": 1932 }, { "epoch": 0.24589746851545605, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.078327178955078, "learning_rate": 8.189910979228485e-07, "loss": 0.4216, "mean_token_accuracy": 0.8608570098876953, "num_tokens": 73907703.0, "step": 1933 }, { "epoch": 0.24602467879404655, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.09131145477295, "learning_rate": 8.194150063586265e-07, "loss": 0.5162, "mean_token_accuracy": 0.8367003202438354, "num_tokens": 73945449.0, "step": 1934 }, { "epoch": 0.24615188907263708, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.19759750366211, "learning_rate": 8.198389147944043e-07, "loss": 0.5036, "mean_token_accuracy": 0.8338149785995483, "num_tokens": 73982461.0, "step": 1935 }, { "epoch": 0.24627909935122758, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.03532886505127, "learning_rate": 8.202628232301823e-07, "loss": 0.4765, "mean_token_accuracy": 0.8494622111320496, "num_tokens": 74018252.0, "step": 1936 }, { "epoch": 0.24640630962981808, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.083036422729492, "learning_rate": 8.206867316659601e-07, "loss": 0.4975, "mean_token_accuracy": 0.8375999927520752, "num_tokens": 74065843.0, "step": 1937 }, { "epoch": 0.2465335199084086, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.081803321838379, "learning_rate": 8.21110640101738e-07, "loss": 0.3884, "mean_token_accuracy": 0.8692976236343384, "num_tokens": 74106389.0, "step": 1938 }, { "epoch": 0.2466607301869991, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.105204582214355, "learning_rate": 8.215345485375159e-07, "loss": 0.5081, "mean_token_accuracy": 0.8364065289497375, "num_tokens": 74150786.0, "step": 1939 }, { "epoch": 0.2467879404655896, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.110336303710938, "learning_rate": 8.219584569732937e-07, "loss": 0.4783, "mean_token_accuracy": 0.8413776159286499, "num_tokens": 74190474.0, "step": 1940 }, { "epoch": 0.24691515074418013, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.093400955200195, "learning_rate": 8.223823654090715e-07, "loss": 0.4286, "mean_token_accuracy": 0.8597997426986694, "num_tokens": 74226437.0, "step": 1941 }, { "epoch": 0.24704236102277063, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.135760307312012, "learning_rate": 8.228062738448495e-07, "loss": 0.4567, "mean_token_accuracy": 0.8488237857818604, "num_tokens": 74267292.0, "step": 1942 }, { "epoch": 0.24716957130136116, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 8.10847282409668, "learning_rate": 8.232301822806273e-07, "loss": 0.4822, "mean_token_accuracy": 0.8434813022613525, "num_tokens": 74306738.0, "step": 1943 }, { "epoch": 0.24729678157995166, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.185600280761719, "learning_rate": 8.236540907164053e-07, "loss": 0.4708, "mean_token_accuracy": 0.8494529724121094, "num_tokens": 74343151.0, "step": 1944 }, { "epoch": 0.24742399185854216, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.12606143951416, "learning_rate": 8.240779991521831e-07, "loss": 0.4482, "mean_token_accuracy": 0.8526180982589722, "num_tokens": 74382755.0, "step": 1945 }, { "epoch": 0.2475512021371327, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.115909576416016, "learning_rate": 8.24501907587961e-07, "loss": 0.4843, "mean_token_accuracy": 0.8433132171630859, "num_tokens": 74426129.0, "step": 1946 }, { "epoch": 0.2476784124157232, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.196830749511719, "learning_rate": 8.249258160237388e-07, "loss": 0.4955, "mean_token_accuracy": 0.8444429636001587, "num_tokens": 74460904.0, "step": 1947 }, { "epoch": 0.2478056226943137, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.220181465148926, "learning_rate": 8.253497244595167e-07, "loss": 0.4849, "mean_token_accuracy": 0.8433274626731873, "num_tokens": 74497993.0, "step": 1948 }, { "epoch": 0.24793283297290422, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.124897003173828, "learning_rate": 8.257736328952945e-07, "loss": 0.5032, "mean_token_accuracy": 0.8378598690032959, "num_tokens": 74541108.0, "step": 1949 }, { "epoch": 0.24806004325149472, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.116312980651855, "learning_rate": 8.261975413310725e-07, "loss": 0.4726, "mean_token_accuracy": 0.8478400707244873, "num_tokens": 74579389.0, "step": 1950 }, { "epoch": 0.24818725353008522, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.159470558166504, "learning_rate": 8.266214497668503e-07, "loss": 0.4262, "mean_token_accuracy": 0.8631418347358704, "num_tokens": 74621546.0, "step": 1951 }, { "epoch": 0.24831446380867575, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.180047035217285, "learning_rate": 8.270453582026283e-07, "loss": 0.5062, "mean_token_accuracy": 0.8346855640411377, "num_tokens": 74660178.0, "step": 1952 }, { "epoch": 0.24844167408726625, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.173078536987305, "learning_rate": 8.274692666384061e-07, "loss": 0.4524, "mean_token_accuracy": 0.8525577783584595, "num_tokens": 74693867.0, "step": 1953 }, { "epoch": 0.24856888436585675, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.156720161437988, "learning_rate": 8.27893175074184e-07, "loss": 0.4658, "mean_token_accuracy": 0.8473145365715027, "num_tokens": 74727561.0, "step": 1954 }, { "epoch": 0.24869609464444728, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.148475646972656, "learning_rate": 8.283170835099618e-07, "loss": 0.4687, "mean_token_accuracy": 0.8490105867385864, "num_tokens": 74763719.0, "step": 1955 }, { "epoch": 0.24882330492303778, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.159041404724121, "learning_rate": 8.287409919457396e-07, "loss": 0.4329, "mean_token_accuracy": 0.8580945730209351, "num_tokens": 74802666.0, "step": 1956 }, { "epoch": 0.24895051520162828, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.17724323272705, "learning_rate": 8.291649003815175e-07, "loss": 0.4836, "mean_token_accuracy": 0.8444124460220337, "num_tokens": 74840307.0, "step": 1957 }, { "epoch": 0.2490777254802188, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.126241683959961, "learning_rate": 8.295888088172954e-07, "loss": 0.495, "mean_token_accuracy": 0.842381477355957, "num_tokens": 74880181.0, "step": 1958 }, { "epoch": 0.2492049357588093, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.145944595336914, "learning_rate": 8.300127172530733e-07, "loss": 0.4028, "mean_token_accuracy": 0.8662599325180054, "num_tokens": 74917148.0, "step": 1959 }, { "epoch": 0.2493321460373998, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.17155647277832, "learning_rate": 8.304366256888512e-07, "loss": 0.4403, "mean_token_accuracy": 0.8542469143867493, "num_tokens": 74953124.0, "step": 1960 }, { "epoch": 0.24945935631599034, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.192473411560059, "learning_rate": 8.308605341246291e-07, "loss": 0.4533, "mean_token_accuracy": 0.8513895869255066, "num_tokens": 74989726.0, "step": 1961 }, { "epoch": 0.24958656659458084, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.14447021484375, "learning_rate": 8.312844425604068e-07, "loss": 0.455, "mean_token_accuracy": 0.8580097556114197, "num_tokens": 75024849.0, "step": 1962 }, { "epoch": 0.24971377687317134, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.189835548400879, "learning_rate": 8.317083509961848e-07, "loss": 0.4821, "mean_token_accuracy": 0.8460061550140381, "num_tokens": 75066464.0, "step": 1963 }, { "epoch": 0.24984098715176187, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.118042945861816, "learning_rate": 8.321322594319626e-07, "loss": 0.434, "mean_token_accuracy": 0.8570204973220825, "num_tokens": 75106829.0, "step": 1964 }, { "epoch": 0.24996819743035237, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.229778289794922, "learning_rate": 8.325561678677405e-07, "loss": 0.4881, "mean_token_accuracy": 0.8429280519485474, "num_tokens": 75148227.0, "step": 1965 }, { "epoch": 0.2500954077089429, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.141191482543945, "learning_rate": 8.329800763035184e-07, "loss": 0.501, "mean_token_accuracy": 0.8368287086486816, "num_tokens": 75188719.0, "step": 1966 }, { "epoch": 0.25022261798753337, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.156000137329102, "learning_rate": 8.334039847392963e-07, "loss": 0.4516, "mean_token_accuracy": 0.8508450984954834, "num_tokens": 75222260.0, "step": 1967 }, { "epoch": 0.2503498282661239, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.17175006866455, "learning_rate": 8.338278931750742e-07, "loss": 0.427, "mean_token_accuracy": 0.860411524772644, "num_tokens": 75255489.0, "step": 1968 }, { "epoch": 0.2504770385447144, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.154865264892578, "learning_rate": 8.342518016108521e-07, "loss": 0.4528, "mean_token_accuracy": 0.8478591442108154, "num_tokens": 75290548.0, "step": 1969 }, { "epoch": 0.2506042488233049, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.209209442138672, "learning_rate": 8.346757100466298e-07, "loss": 0.4536, "mean_token_accuracy": 0.852687418460846, "num_tokens": 75323912.0, "step": 1970 }, { "epoch": 0.2507314591018954, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.12789535522461, "learning_rate": 8.350996184824078e-07, "loss": 0.4941, "mean_token_accuracy": 0.8374795913696289, "num_tokens": 75361206.0, "step": 1971 }, { "epoch": 0.25085866938048595, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.200345039367676, "learning_rate": 8.355235269181856e-07, "loss": 0.4748, "mean_token_accuracy": 0.8452510237693787, "num_tokens": 75397292.0, "step": 1972 }, { "epoch": 0.2509858796590764, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.227422714233398, "learning_rate": 8.359474353539635e-07, "loss": 0.4343, "mean_token_accuracy": 0.8593279123306274, "num_tokens": 75437616.0, "step": 1973 }, { "epoch": 0.25111308993766696, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.185281753540039, "learning_rate": 8.363713437897414e-07, "loss": 0.5244, "mean_token_accuracy": 0.8327155709266663, "num_tokens": 75486172.0, "step": 1974 }, { "epoch": 0.2512403002162575, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.153949737548828, "learning_rate": 8.367952522255193e-07, "loss": 0.4592, "mean_token_accuracy": 0.8476098775863647, "num_tokens": 75526637.0, "step": 1975 }, { "epoch": 0.25136751049484796, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.244026184082031, "learning_rate": 8.372191606612972e-07, "loss": 0.4611, "mean_token_accuracy": 0.8515615463256836, "num_tokens": 75564632.0, "step": 1976 }, { "epoch": 0.2514947207734385, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.206006050109863, "learning_rate": 8.376430690970749e-07, "loss": 0.4289, "mean_token_accuracy": 0.861152708530426, "num_tokens": 75595553.0, "step": 1977 }, { "epoch": 0.251621931052029, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.131875038146973, "learning_rate": 8.380669775328528e-07, "loss": 0.4235, "mean_token_accuracy": 0.8631882667541504, "num_tokens": 75635847.0, "step": 1978 }, { "epoch": 0.25174914133061954, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.202776908874512, "learning_rate": 8.384908859686307e-07, "loss": 0.4216, "mean_token_accuracy": 0.8638254404067993, "num_tokens": 75677403.0, "step": 1979 }, { "epoch": 0.25187635160921, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.130854606628418, "learning_rate": 8.389147944044086e-07, "loss": 0.4627, "mean_token_accuracy": 0.849134087562561, "num_tokens": 75720703.0, "step": 1980 }, { "epoch": 0.25200356188780054, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.181777000427246, "learning_rate": 8.393387028401864e-07, "loss": 0.4869, "mean_token_accuracy": 0.8432750701904297, "num_tokens": 75758000.0, "step": 1981 }, { "epoch": 0.25213077216639107, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.24954605102539, "learning_rate": 8.397626112759644e-07, "loss": 0.5135, "mean_token_accuracy": 0.8351296186447144, "num_tokens": 75791113.0, "step": 1982 }, { "epoch": 0.25225798244498154, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.257591247558594, "learning_rate": 8.401865197117422e-07, "loss": 0.4594, "mean_token_accuracy": 0.8510538339614868, "num_tokens": 75821733.0, "step": 1983 }, { "epoch": 0.25238519272357207, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.167801856994629, "learning_rate": 8.406104281475202e-07, "loss": 0.4379, "mean_token_accuracy": 0.8552919626235962, "num_tokens": 75861404.0, "step": 1984 }, { "epoch": 0.2525124030021626, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.167705535888672, "learning_rate": 8.410343365832979e-07, "loss": 0.4387, "mean_token_accuracy": 0.8558186888694763, "num_tokens": 75903059.0, "step": 1985 }, { "epoch": 0.2526396132807531, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.192361831665039, "learning_rate": 8.414582450190758e-07, "loss": 0.451, "mean_token_accuracy": 0.8537959456443787, "num_tokens": 75947125.0, "step": 1986 }, { "epoch": 0.2527668235593436, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.211221694946289, "learning_rate": 8.418821534548537e-07, "loss": 0.5355, "mean_token_accuracy": 0.828041136264801, "num_tokens": 75987519.0, "step": 1987 }, { "epoch": 0.25289403383793413, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.246759414672852, "learning_rate": 8.423060618906316e-07, "loss": 0.4613, "mean_token_accuracy": 0.8503937721252441, "num_tokens": 76027728.0, "step": 1988 }, { "epoch": 0.2530212441165246, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.162484169006348, "learning_rate": 8.427299703264095e-07, "loss": 0.4601, "mean_token_accuracy": 0.8519246578216553, "num_tokens": 76070942.0, "step": 1989 }, { "epoch": 0.25314845439511513, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.195629119873047, "learning_rate": 8.431538787621874e-07, "loss": 0.4856, "mean_token_accuracy": 0.8440224528312683, "num_tokens": 76113330.0, "step": 1990 }, { "epoch": 0.25327566467370566, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.18897533416748, "learning_rate": 8.435777871979652e-07, "loss": 0.4391, "mean_token_accuracy": 0.8571264743804932, "num_tokens": 76153511.0, "step": 1991 }, { "epoch": 0.25340287495229613, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.257387161254883, "learning_rate": 8.440016956337432e-07, "loss": 0.4641, "mean_token_accuracy": 0.8508033752441406, "num_tokens": 76191505.0, "step": 1992 }, { "epoch": 0.25353008523088666, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.24412727355957, "learning_rate": 8.444256040695209e-07, "loss": 0.48, "mean_token_accuracy": 0.8467576503753662, "num_tokens": 76232479.0, "step": 1993 }, { "epoch": 0.2536572955094772, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.194162368774414, "learning_rate": 8.448495125052988e-07, "loss": 0.4734, "mean_token_accuracy": 0.8483437299728394, "num_tokens": 76270192.0, "step": 1994 }, { "epoch": 0.25378450578806766, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.258111953735352, "learning_rate": 8.452734209410767e-07, "loss": 0.4139, "mean_token_accuracy": 0.8652498126029968, "num_tokens": 76303981.0, "step": 1995 }, { "epoch": 0.2539117160666582, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.199419021606445, "learning_rate": 8.456973293768545e-07, "loss": 0.4381, "mean_token_accuracy": 0.8553071022033691, "num_tokens": 76338496.0, "step": 1996 }, { "epoch": 0.2540389263452487, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.233875274658203, "learning_rate": 8.461212378126325e-07, "loss": 0.4921, "mean_token_accuracy": 0.8418352603912354, "num_tokens": 76377904.0, "step": 1997 }, { "epoch": 0.2541661366238392, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.270254135131836, "learning_rate": 8.465451462484103e-07, "loss": 0.5151, "mean_token_accuracy": 0.8362829089164734, "num_tokens": 76413905.0, "step": 1998 }, { "epoch": 0.2542933469024297, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.24566650390625, "learning_rate": 8.469690546841882e-07, "loss": 0.4645, "mean_token_accuracy": 0.8486731052398682, "num_tokens": 76451591.0, "step": 1999 }, { "epoch": 0.25442055718102025, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.253108978271484, "learning_rate": 8.47392963119966e-07, "loss": 0.4442, "mean_token_accuracy": 0.8553458452224731, "num_tokens": 76486191.0, "step": 2000 }, { "epoch": 0.2545477674596107, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.278841018676758, "learning_rate": 8.478168715557439e-07, "loss": 0.464, "mean_token_accuracy": 0.8513522148132324, "num_tokens": 76523317.0, "step": 2001 }, { "epoch": 0.25467497773820125, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.274680137634277, "learning_rate": 8.482407799915217e-07, "loss": 0.4707, "mean_token_accuracy": 0.8471559286117554, "num_tokens": 76562661.0, "step": 2002 }, { "epoch": 0.2548021880167918, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.227750778198242, "learning_rate": 8.486646884272997e-07, "loss": 0.4649, "mean_token_accuracy": 0.8473291397094727, "num_tokens": 76600914.0, "step": 2003 }, { "epoch": 0.25492939829538225, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.287372589111328, "learning_rate": 8.490885968630775e-07, "loss": 0.4849, "mean_token_accuracy": 0.8414118885993958, "num_tokens": 76636310.0, "step": 2004 }, { "epoch": 0.2550566085739728, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.304034233093262, "learning_rate": 8.495125052988555e-07, "loss": 0.4696, "mean_token_accuracy": 0.8508148193359375, "num_tokens": 76680889.0, "step": 2005 }, { "epoch": 0.2551838188525633, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.258349418640137, "learning_rate": 8.499364137346333e-07, "loss": 0.4621, "mean_token_accuracy": 0.8542088270187378, "num_tokens": 76722469.0, "step": 2006 }, { "epoch": 0.2553110291311538, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.28075885772705, "learning_rate": 8.503603221704112e-07, "loss": 0.4972, "mean_token_accuracy": 0.839238166809082, "num_tokens": 76757668.0, "step": 2007 }, { "epoch": 0.2554382394097443, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.231954574584961, "learning_rate": 8.50784230606189e-07, "loss": 0.4585, "mean_token_accuracy": 0.8520522117614746, "num_tokens": 76795607.0, "step": 2008 }, { "epoch": 0.25556544968833483, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.325827598571777, "learning_rate": 8.512081390419669e-07, "loss": 0.4871, "mean_token_accuracy": 0.8496173620223999, "num_tokens": 76837161.0, "step": 2009 }, { "epoch": 0.2556926599669253, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.23212718963623, "learning_rate": 8.516320474777447e-07, "loss": 0.4664, "mean_token_accuracy": 0.8496169447898865, "num_tokens": 76872198.0, "step": 2010 }, { "epoch": 0.25581987024551583, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.224824905395508, "learning_rate": 8.520559559135227e-07, "loss": 0.4317, "mean_token_accuracy": 0.8624874353408813, "num_tokens": 76908976.0, "step": 2011 }, { "epoch": 0.25594708052410636, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.320119857788086, "learning_rate": 8.524798643493005e-07, "loss": 0.4829, "mean_token_accuracy": 0.8483908772468567, "num_tokens": 76947298.0, "step": 2012 }, { "epoch": 0.25607429080269684, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.204353332519531, "learning_rate": 8.529037727850785e-07, "loss": 0.4382, "mean_token_accuracy": 0.8554422855377197, "num_tokens": 76985112.0, "step": 2013 }, { "epoch": 0.25620150108128736, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.334572792053223, "learning_rate": 8.533276812208563e-07, "loss": 0.4117, "mean_token_accuracy": 0.8643877506256104, "num_tokens": 77020309.0, "step": 2014 }, { "epoch": 0.2563287113598779, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.263723373413086, "learning_rate": 8.53751589656634e-07, "loss": 0.4514, "mean_token_accuracy": 0.8506179451942444, "num_tokens": 77063245.0, "step": 2015 }, { "epoch": 0.25645592163846836, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.30891227722168, "learning_rate": 8.54175498092412e-07, "loss": 0.4486, "mean_token_accuracy": 0.8538740873336792, "num_tokens": 77096791.0, "step": 2016 }, { "epoch": 0.2565831319170589, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.318470001220703, "learning_rate": 8.545994065281898e-07, "loss": 0.4282, "mean_token_accuracy": 0.8611332178115845, "num_tokens": 77129281.0, "step": 2017 }, { "epoch": 0.2567103421956494, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.278250694274902, "learning_rate": 8.550233149639677e-07, "loss": 0.5154, "mean_token_accuracy": 0.8319746255874634, "num_tokens": 77168785.0, "step": 2018 }, { "epoch": 0.2568375524742399, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.281807899475098, "learning_rate": 8.554472233997456e-07, "loss": 0.4831, "mean_token_accuracy": 0.8414247035980225, "num_tokens": 77208360.0, "step": 2019 }, { "epoch": 0.2569647627528304, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.317639350891113, "learning_rate": 8.558711318355235e-07, "loss": 0.5157, "mean_token_accuracy": 0.8321080207824707, "num_tokens": 77245539.0, "step": 2020 }, { "epoch": 0.25709197303142095, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.299196243286133, "learning_rate": 8.562950402713014e-07, "loss": 0.4181, "mean_token_accuracy": 0.8688679933547974, "num_tokens": 77279118.0, "step": 2021 }, { "epoch": 0.2572191833100114, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.315873146057129, "learning_rate": 8.567189487070793e-07, "loss": 0.5053, "mean_token_accuracy": 0.8399206399917603, "num_tokens": 77316259.0, "step": 2022 }, { "epoch": 0.25734639358860195, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.26960563659668, "learning_rate": 8.57142857142857e-07, "loss": 0.5227, "mean_token_accuracy": 0.8324644565582275, "num_tokens": 77359712.0, "step": 2023 }, { "epoch": 0.2574736038671925, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.301711082458496, "learning_rate": 8.57566765578635e-07, "loss": 0.4266, "mean_token_accuracy": 0.8641119003295898, "num_tokens": 77402269.0, "step": 2024 }, { "epoch": 0.25760081414578295, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.261146545410156, "learning_rate": 8.579906740144128e-07, "loss": 0.4554, "mean_token_accuracy": 0.8535125255584717, "num_tokens": 77445791.0, "step": 2025 }, { "epoch": 0.2577280244243735, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.289194107055664, "learning_rate": 8.584145824501907e-07, "loss": 0.4886, "mean_token_accuracy": 0.8398106098175049, "num_tokens": 77485376.0, "step": 2026 }, { "epoch": 0.257855234702964, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.30396556854248, "learning_rate": 8.588384908859686e-07, "loss": 0.485, "mean_token_accuracy": 0.8466249108314514, "num_tokens": 77524344.0, "step": 2027 }, { "epoch": 0.25798244498155454, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.277426719665527, "learning_rate": 8.592623993217465e-07, "loss": 0.4783, "mean_token_accuracy": 0.8476002812385559, "num_tokens": 77562746.0, "step": 2028 }, { "epoch": 0.258109655260145, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.351402282714844, "learning_rate": 8.596863077575244e-07, "loss": 0.501, "mean_token_accuracy": 0.8416749238967896, "num_tokens": 77601261.0, "step": 2029 }, { "epoch": 0.25823686553873554, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.355213165283203, "learning_rate": 8.601102161933023e-07, "loss": 0.4816, "mean_token_accuracy": 0.8449110984802246, "num_tokens": 77640054.0, "step": 2030 }, { "epoch": 0.25836407581732607, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.33460807800293, "learning_rate": 8.6053412462908e-07, "loss": 0.4558, "mean_token_accuracy": 0.8524748086929321, "num_tokens": 77680098.0, "step": 2031 }, { "epoch": 0.25849128609591654, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.38382339477539, "learning_rate": 8.60958033064858e-07, "loss": 0.4498, "mean_token_accuracy": 0.8562577366828918, "num_tokens": 77712517.0, "step": 2032 }, { "epoch": 0.25861849637450707, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.33108139038086, "learning_rate": 8.613819415006358e-07, "loss": 0.5119, "mean_token_accuracy": 0.8414668440818787, "num_tokens": 77751335.0, "step": 2033 }, { "epoch": 0.2587457066530976, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.401446342468262, "learning_rate": 8.618058499364137e-07, "loss": 0.4993, "mean_token_accuracy": 0.8390690088272095, "num_tokens": 77792300.0, "step": 2034 }, { "epoch": 0.25887291693168807, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.312236785888672, "learning_rate": 8.622297583721916e-07, "loss": 0.4577, "mean_token_accuracy": 0.8550906181335449, "num_tokens": 77836391.0, "step": 2035 }, { "epoch": 0.2590001272102786, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.343389511108398, "learning_rate": 8.626536668079695e-07, "loss": 0.4683, "mean_token_accuracy": 0.8470888137817383, "num_tokens": 77871600.0, "step": 2036 }, { "epoch": 0.2591273374888691, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.407554626464844, "learning_rate": 8.630775752437474e-07, "loss": 0.4567, "mean_token_accuracy": 0.8525091409683228, "num_tokens": 77904623.0, "step": 2037 }, { "epoch": 0.2592545477674596, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.406428337097168, "learning_rate": 8.635014836795251e-07, "loss": 0.4952, "mean_token_accuracy": 0.841690182685852, "num_tokens": 77941145.0, "step": 2038 }, { "epoch": 0.2593817580460501, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.365666389465332, "learning_rate": 8.63925392115303e-07, "loss": 0.4331, "mean_token_accuracy": 0.8567731380462646, "num_tokens": 77975087.0, "step": 2039 }, { "epoch": 0.25950896832464065, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.429258346557617, "learning_rate": 8.643493005510809e-07, "loss": 0.4402, "mean_token_accuracy": 0.8559155464172363, "num_tokens": 78009526.0, "step": 2040 }, { "epoch": 0.2596361786032311, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.313314437866211, "learning_rate": 8.647732089868588e-07, "loss": 0.5175, "mean_token_accuracy": 0.8327131271362305, "num_tokens": 78050918.0, "step": 2041 }, { "epoch": 0.25976338888182166, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.404251098632812, "learning_rate": 8.651971174226366e-07, "loss": 0.4348, "mean_token_accuracy": 0.8590621948242188, "num_tokens": 78085547.0, "step": 2042 }, { "epoch": 0.2598905991604122, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.396203994750977, "learning_rate": 8.656210258584146e-07, "loss": 0.5035, "mean_token_accuracy": 0.8360130786895752, "num_tokens": 78122916.0, "step": 2043 }, { "epoch": 0.26001780943900266, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.386323928833008, "learning_rate": 8.660449342941924e-07, "loss": 0.4964, "mean_token_accuracy": 0.8372600078582764, "num_tokens": 78161800.0, "step": 2044 }, { "epoch": 0.2601450197175932, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.377896308898926, "learning_rate": 8.664688427299704e-07, "loss": 0.475, "mean_token_accuracy": 0.8462398052215576, "num_tokens": 78199961.0, "step": 2045 }, { "epoch": 0.2602722299961837, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.435422897338867, "learning_rate": 8.668927511657481e-07, "loss": 0.4333, "mean_token_accuracy": 0.8564614057540894, "num_tokens": 78231332.0, "step": 2046 }, { "epoch": 0.2603994402747742, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.405617713928223, "learning_rate": 8.67316659601526e-07, "loss": 0.4895, "mean_token_accuracy": 0.840456485748291, "num_tokens": 78265054.0, "step": 2047 }, { "epoch": 0.2605266505533647, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.392800331115723, "learning_rate": 8.677405680373039e-07, "loss": 0.4462, "mean_token_accuracy": 0.8574856519699097, "num_tokens": 78305445.0, "step": 2048 }, { "epoch": 0.26065386083195524, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.428120613098145, "learning_rate": 8.681644764730818e-07, "loss": 0.4747, "mean_token_accuracy": 0.8460973501205444, "num_tokens": 78339174.0, "step": 2049 }, { "epoch": 0.2607810711105457, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.326119422912598, "learning_rate": 8.685883849088596e-07, "loss": 0.4175, "mean_token_accuracy": 0.8635805249214172, "num_tokens": 78377630.0, "step": 2050 }, { "epoch": 0.26090828138913624, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.356456756591797, "learning_rate": 8.690122933446376e-07, "loss": 0.4746, "mean_token_accuracy": 0.8497322797775269, "num_tokens": 78419908.0, "step": 2051 }, { "epoch": 0.26103549166772677, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.382433891296387, "learning_rate": 8.694362017804154e-07, "loss": 0.4409, "mean_token_accuracy": 0.8582432270050049, "num_tokens": 78459561.0, "step": 2052 }, { "epoch": 0.26116270194631724, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.345514297485352, "learning_rate": 8.698601102161933e-07, "loss": 0.4723, "mean_token_accuracy": 0.8464169502258301, "num_tokens": 78502500.0, "step": 2053 }, { "epoch": 0.26128991222490777, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.385196685791016, "learning_rate": 8.702840186519711e-07, "loss": 0.4299, "mean_token_accuracy": 0.8603520393371582, "num_tokens": 78545411.0, "step": 2054 }, { "epoch": 0.2614171225034983, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.383088111877441, "learning_rate": 8.70707927087749e-07, "loss": 0.4296, "mean_token_accuracy": 0.8604910373687744, "num_tokens": 78585878.0, "step": 2055 }, { "epoch": 0.2615443327820888, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.436989784240723, "learning_rate": 8.711318355235269e-07, "loss": 0.5186, "mean_token_accuracy": 0.8308005332946777, "num_tokens": 78621233.0, "step": 2056 }, { "epoch": 0.2616715430606793, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.40007495880127, "learning_rate": 8.715557439593047e-07, "loss": 0.4382, "mean_token_accuracy": 0.8555337190628052, "num_tokens": 78654644.0, "step": 2057 }, { "epoch": 0.26179875333926983, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.400352478027344, "learning_rate": 8.719796523950826e-07, "loss": 0.515, "mean_token_accuracy": 0.8321032524108887, "num_tokens": 78692851.0, "step": 2058 }, { "epoch": 0.2619259636178603, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.399434089660645, "learning_rate": 8.724035608308605e-07, "loss": 0.409, "mean_token_accuracy": 0.8663186430931091, "num_tokens": 78725329.0, "step": 2059 }, { "epoch": 0.26205317389645083, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.374418258666992, "learning_rate": 8.728274692666384e-07, "loss": 0.4656, "mean_token_accuracy": 0.8502508401870728, "num_tokens": 78766645.0, "step": 2060 }, { "epoch": 0.26218038417504136, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.408496856689453, "learning_rate": 8.732513777024162e-07, "loss": 0.4605, "mean_token_accuracy": 0.8523497581481934, "num_tokens": 78801576.0, "step": 2061 }, { "epoch": 0.26230759445363183, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.372028350830078, "learning_rate": 8.736752861381941e-07, "loss": 0.5035, "mean_token_accuracy": 0.8368909358978271, "num_tokens": 78845577.0, "step": 2062 }, { "epoch": 0.26243480473222236, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.41567325592041, "learning_rate": 8.740991945739719e-07, "loss": 0.4688, "mean_token_accuracy": 0.8488438129425049, "num_tokens": 78884229.0, "step": 2063 }, { "epoch": 0.2625620150108129, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.379924774169922, "learning_rate": 8.745231030097499e-07, "loss": 0.4558, "mean_token_accuracy": 0.852318525314331, "num_tokens": 78926948.0, "step": 2064 }, { "epoch": 0.26268922528940336, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.43273639678955, "learning_rate": 8.749470114455277e-07, "loss": 0.4323, "mean_token_accuracy": 0.8583897352218628, "num_tokens": 78963582.0, "step": 2065 }, { "epoch": 0.2628164355679939, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.419231414794922, "learning_rate": 8.753709198813056e-07, "loss": 0.5392, "mean_token_accuracy": 0.8267940878868103, "num_tokens": 79007345.0, "step": 2066 }, { "epoch": 0.2629436458465844, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.422785758972168, "learning_rate": 8.757948283170835e-07, "loss": 0.4368, "mean_token_accuracy": 0.8604376316070557, "num_tokens": 79049505.0, "step": 2067 }, { "epoch": 0.2630708561251749, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.990795135498047e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.445603370666504, "learning_rate": 8.762187367528613e-07, "loss": 0.4317, "mean_token_accuracy": 0.8564521670341492, "num_tokens": 79087419.0, "step": 2068 }, { "epoch": 0.2631980664037654, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.990795135498047e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.423041343688965, "learning_rate": 8.766426451886392e-07, "loss": 0.4792, "mean_token_accuracy": 0.8439714908599854, "num_tokens": 79131444.0, "step": 2069 }, { "epoch": 0.26332527668235595, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.990795135498047e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.38815975189209, "learning_rate": 8.770665536244171e-07, "loss": 0.5058, "mean_token_accuracy": 0.8403980731964111, "num_tokens": 79171976.0, "step": 2070 }, { "epoch": 0.2634524869609464, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.990795135498047e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.453375816345215, "learning_rate": 8.774904620601949e-07, "loss": 0.4349, "mean_token_accuracy": 0.8569670915603638, "num_tokens": 79206273.0, "step": 2071 }, { "epoch": 0.26357969723953695, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.990795135498047e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.453447341918945, "learning_rate": 8.779143704959729e-07, "loss": 0.4358, "mean_token_accuracy": 0.8571427464485168, "num_tokens": 79245564.0, "step": 2072 }, { "epoch": 0.2637069075181275, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.990795135498047e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.429100036621094, "learning_rate": 8.783382789317507e-07, "loss": 0.4671, "mean_token_accuracy": 0.8487162590026855, "num_tokens": 79281857.0, "step": 2073 }, { "epoch": 0.26383411779671795, "ewc_loss": 0.01025390625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.478113174438477, "learning_rate": 8.787621873675286e-07, "loss": 0.5004, "mean_token_accuracy": 0.8348530530929565, "num_tokens": 79312482.0, "step": 2074 }, { "epoch": 0.2639613280753085, "ewc_loss": 0.01025390625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.433320999145508, "learning_rate": 8.791860958033065e-07, "loss": 0.4396, "mean_token_accuracy": 0.8575518131256104, "num_tokens": 79348157.0, "step": 2075 }, { "epoch": 0.264088538353899, "ewc_loss": 0.01025390625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.402637481689453, "learning_rate": 8.796100042390842e-07, "loss": 0.4261, "mean_token_accuracy": 0.8621498942375183, "num_tokens": 79387540.0, "step": 2076 }, { "epoch": 0.2642157486324895, "ewc_loss": 0.01025390625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.505059242248535, "learning_rate": 8.800339126748622e-07, "loss": 0.4436, "mean_token_accuracy": 0.8545414805412292, "num_tokens": 79421424.0, "step": 2077 }, { "epoch": 0.26434295891108, "ewc_loss": 0.01025390625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.459012985229492, "learning_rate": 8.8045782111064e-07, "loss": 0.5133, "mean_token_accuracy": 0.8338460922241211, "num_tokens": 79458733.0, "step": 2078 }, { "epoch": 0.26447016918967053, "ewc_loss": 0.01025390625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.466694831848145, "learning_rate": 8.808817295464179e-07, "loss": 0.4656, "mean_token_accuracy": 0.8505342602729797, "num_tokens": 79495394.0, "step": 2079 }, { "epoch": 0.26459737946826106, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.506178855895996, "learning_rate": 8.813056379821958e-07, "loss": 0.4363, "mean_token_accuracy": 0.8598596453666687, "num_tokens": 79529799.0, "step": 2080 }, { "epoch": 0.26472458974685154, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.462308883666992, "learning_rate": 8.817295464179737e-07, "loss": 0.4624, "mean_token_accuracy": 0.8499364852905273, "num_tokens": 79564664.0, "step": 2081 }, { "epoch": 0.26485180002544206, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.504250526428223, "learning_rate": 8.821534548537515e-07, "loss": 0.5401, "mean_token_accuracy": 0.8241896629333496, "num_tokens": 79606104.0, "step": 2082 }, { "epoch": 0.2649790103040326, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.451614379882812, "learning_rate": 8.825773632895295e-07, "loss": 0.4362, "mean_token_accuracy": 0.8589428663253784, "num_tokens": 79647688.0, "step": 2083 }, { "epoch": 0.26510622058262306, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.548066139221191, "learning_rate": 8.830012717253072e-07, "loss": 0.4532, "mean_token_accuracy": 0.8554776906967163, "num_tokens": 79685500.0, "step": 2084 }, { "epoch": 0.2652334308612136, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.478649139404297, "learning_rate": 8.834251801610852e-07, "loss": 0.482, "mean_token_accuracy": 0.8429281711578369, "num_tokens": 79729912.0, "step": 2085 }, { "epoch": 0.2653606411398041, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.491031646728516, "learning_rate": 8.83849088596863e-07, "loss": 0.4383, "mean_token_accuracy": 0.8578357696533203, "num_tokens": 79767487.0, "step": 2086 }, { "epoch": 0.2654878514183946, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.48005485534668, "learning_rate": 8.842729970326409e-07, "loss": 0.4189, "mean_token_accuracy": 0.8657627701759338, "num_tokens": 79804802.0, "step": 2087 }, { "epoch": 0.2656150616969851, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.47192096710205, "learning_rate": 8.846969054684188e-07, "loss": 0.5362, "mean_token_accuracy": 0.8300421237945557, "num_tokens": 79845685.0, "step": 2088 }, { "epoch": 0.26574227197557565, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.493416786193848, "learning_rate": 8.851208139041967e-07, "loss": 0.4253, "mean_token_accuracy": 0.8603179454803467, "num_tokens": 79884692.0, "step": 2089 }, { "epoch": 0.2658694822541661, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.506133079528809, "learning_rate": 8.855447223399745e-07, "loss": 0.4846, "mean_token_accuracy": 0.8410308361053467, "num_tokens": 79919780.0, "step": 2090 }, { "epoch": 0.26599669253275665, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.539575576782227, "learning_rate": 8.859686307757524e-07, "loss": 0.5103, "mean_token_accuracy": 0.8351059556007385, "num_tokens": 79958317.0, "step": 2091 }, { "epoch": 0.2661239028113472, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.461048126220703, "learning_rate": 8.863925392115302e-07, "loss": 0.4436, "mean_token_accuracy": 0.8548001050949097, "num_tokens": 79994126.0, "step": 2092 }, { "epoch": 0.26625111308993765, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.54304027557373, "learning_rate": 8.868164476473082e-07, "loss": 0.4131, "mean_token_accuracy": 0.8655351400375366, "num_tokens": 80033157.0, "step": 2093 }, { "epoch": 0.2663783233685282, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.477413177490234, "learning_rate": 8.87240356083086e-07, "loss": 0.442, "mean_token_accuracy": 0.856958270072937, "num_tokens": 80073354.0, "step": 2094 }, { "epoch": 0.2665055336471187, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.52357006072998, "learning_rate": 8.876642645188639e-07, "loss": 0.5025, "mean_token_accuracy": 0.8369015455245972, "num_tokens": 80111819.0, "step": 2095 }, { "epoch": 0.2666327439257092, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.52033805847168, "learning_rate": 8.880881729546418e-07, "loss": 0.4166, "mean_token_accuracy": 0.8637397885322571, "num_tokens": 80147951.0, "step": 2096 }, { "epoch": 0.2667599542042997, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.51229190826416, "learning_rate": 8.885120813904197e-07, "loss": 0.4315, "mean_token_accuracy": 0.8582104444503784, "num_tokens": 80191319.0, "step": 2097 }, { "epoch": 0.26688716448289024, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.571447372436523, "learning_rate": 8.889359898261976e-07, "loss": 0.518, "mean_token_accuracy": 0.8361396789550781, "num_tokens": 80226300.0, "step": 2098 }, { "epoch": 0.2670143747614807, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.507828712463379, "learning_rate": 8.893598982619753e-07, "loss": 0.4437, "mean_token_accuracy": 0.8557595014572144, "num_tokens": 80265465.0, "step": 2099 }, { "epoch": 0.26714158504007124, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.543033599853516, "learning_rate": 8.897838066977532e-07, "loss": 0.4464, "mean_token_accuracy": 0.8551235198974609, "num_tokens": 80312696.0, "step": 2100 }, { "epoch": 0.26726879531866177, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.553933143615723, "learning_rate": 8.902077151335311e-07, "loss": 0.4665, "mean_token_accuracy": 0.847341775894165, "num_tokens": 80353507.0, "step": 2101 }, { "epoch": 0.26739600559725224, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.551828384399414, "learning_rate": 8.90631623569309e-07, "loss": 0.5492, "mean_token_accuracy": 0.8200649619102478, "num_tokens": 80390795.0, "step": 2102 }, { "epoch": 0.26752321587584277, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.530332565307617, "learning_rate": 8.910555320050868e-07, "loss": 0.4589, "mean_token_accuracy": 0.8514834046363831, "num_tokens": 80427599.0, "step": 2103 }, { "epoch": 0.2676504261544333, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.485198974609375, "learning_rate": 8.914794404408648e-07, "loss": 0.4615, "mean_token_accuracy": 0.8508797287940979, "num_tokens": 80468084.0, "step": 2104 }, { "epoch": 0.26777763643302377, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.545365333557129, "learning_rate": 8.919033488766426e-07, "loss": 0.4057, "mean_token_accuracy": 0.8678247928619385, "num_tokens": 80507218.0, "step": 2105 }, { "epoch": 0.2679048467116143, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.537020683288574, "learning_rate": 8.923272573124204e-07, "loss": 0.4895, "mean_token_accuracy": 0.8428925275802612, "num_tokens": 80541730.0, "step": 2106 }, { "epoch": 0.2680320569902048, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.523205757141113, "learning_rate": 8.927511657481983e-07, "loss": 0.5002, "mean_token_accuracy": 0.83748859167099, "num_tokens": 80584677.0, "step": 2107 }, { "epoch": 0.2681592672687953, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.594012260437012, "learning_rate": 8.931750741839762e-07, "loss": 0.4352, "mean_token_accuracy": 0.860553503036499, "num_tokens": 80616332.0, "step": 2108 }, { "epoch": 0.2682864775473858, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.568090438842773, "learning_rate": 8.935989826197541e-07, "loss": 0.4439, "mean_token_accuracy": 0.8582042455673218, "num_tokens": 80654435.0, "step": 2109 }, { "epoch": 0.26841368782597635, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.504966735839844, "learning_rate": 8.94022891055532e-07, "loss": 0.4208, "mean_token_accuracy": 0.866168737411499, "num_tokens": 80691246.0, "step": 2110 }, { "epoch": 0.2685408981045668, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.570481300354004, "learning_rate": 8.944467994913098e-07, "loss": 0.4719, "mean_token_accuracy": 0.8438059091567993, "num_tokens": 80730510.0, "step": 2111 }, { "epoch": 0.26866810838315736, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.554807662963867, "learning_rate": 8.948707079270878e-07, "loss": 0.3903, "mean_token_accuracy": 0.872146725654602, "num_tokens": 80769369.0, "step": 2112 }, { "epoch": 0.2687953186617479, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.552353858947754, "learning_rate": 8.952946163628656e-07, "loss": 0.4453, "mean_token_accuracy": 0.8545699715614319, "num_tokens": 80802168.0, "step": 2113 }, { "epoch": 0.26892252894033836, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.584298133850098, "learning_rate": 8.957185247986434e-07, "loss": 0.3954, "mean_token_accuracy": 0.8711022734642029, "num_tokens": 80839389.0, "step": 2114 }, { "epoch": 0.2690497392189289, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.548824310302734, "learning_rate": 8.961424332344213e-07, "loss": 0.4506, "mean_token_accuracy": 0.8552147746086121, "num_tokens": 80881139.0, "step": 2115 }, { "epoch": 0.2691769494975194, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.538963317871094, "learning_rate": 8.965663416701992e-07, "loss": 0.4526, "mean_token_accuracy": 0.8513243198394775, "num_tokens": 80919809.0, "step": 2116 }, { "epoch": 0.2693041597761099, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.565673828125, "learning_rate": 8.969902501059771e-07, "loss": 0.4285, "mean_token_accuracy": 0.8627727031707764, "num_tokens": 80956860.0, "step": 2117 }, { "epoch": 0.2694313700547004, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.533601760864258, "learning_rate": 8.97414158541755e-07, "loss": 0.3954, "mean_token_accuracy": 0.8702753782272339, "num_tokens": 80994322.0, "step": 2118 }, { "epoch": 0.26955858033329094, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.558839797973633, "learning_rate": 8.978380669775328e-07, "loss": 0.4787, "mean_token_accuracy": 0.8452726602554321, "num_tokens": 81032792.0, "step": 2119 }, { "epoch": 0.2696857906118814, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.582837104797363, "learning_rate": 8.982619754133107e-07, "loss": 0.4789, "mean_token_accuracy": 0.8457959294319153, "num_tokens": 81070011.0, "step": 2120 }, { "epoch": 0.26981300089047194, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.639252662658691, "learning_rate": 8.986858838490886e-07, "loss": 0.4815, "mean_token_accuracy": 0.8447598814964294, "num_tokens": 81113145.0, "step": 2121 }, { "epoch": 0.26994021116906247, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.466889381408691, "learning_rate": 8.991097922848663e-07, "loss": 0.4493, "mean_token_accuracy": 0.8558201789855957, "num_tokens": 81154175.0, "step": 2122 }, { "epoch": 0.27006742144765294, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.60233211517334, "learning_rate": 8.995337007206443e-07, "loss": 0.4731, "mean_token_accuracy": 0.8486687541007996, "num_tokens": 81193319.0, "step": 2123 }, { "epoch": 0.2701946317262435, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.599145889282227, "learning_rate": 8.999576091564221e-07, "loss": 0.4411, "mean_token_accuracy": 0.856652021408081, "num_tokens": 81229057.0, "step": 2124 }, { "epoch": 0.270321842004834, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.557249069213867, "learning_rate": 9.003815175922001e-07, "loss": 0.4491, "mean_token_accuracy": 0.8544098138809204, "num_tokens": 81262380.0, "step": 2125 }, { "epoch": 0.2704490522834245, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.597082138061523, "learning_rate": 9.008054260279779e-07, "loss": 0.4542, "mean_token_accuracy": 0.8542908430099487, "num_tokens": 81304080.0, "step": 2126 }, { "epoch": 0.270576262562015, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.59877872467041, "learning_rate": 9.012293344637558e-07, "loss": 0.4838, "mean_token_accuracy": 0.8455191254615784, "num_tokens": 81340226.0, "step": 2127 }, { "epoch": 0.27070347284060553, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.599735260009766, "learning_rate": 9.016532428995337e-07, "loss": 0.4306, "mean_token_accuracy": 0.8588836193084717, "num_tokens": 81378046.0, "step": 2128 }, { "epoch": 0.27083068311919606, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.574158668518066, "learning_rate": 9.020771513353115e-07, "loss": 0.4747, "mean_token_accuracy": 0.8460743427276611, "num_tokens": 81419486.0, "step": 2129 }, { "epoch": 0.27095789339778653, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.578628540039062, "learning_rate": 9.025010597710894e-07, "loss": 0.4737, "mean_token_accuracy": 0.8462024331092834, "num_tokens": 81457488.0, "step": 2130 }, { "epoch": 0.27108510367637706, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.562958717346191, "learning_rate": 9.029249682068673e-07, "loss": 0.4428, "mean_token_accuracy": 0.8569331765174866, "num_tokens": 81494812.0, "step": 2131 }, { "epoch": 0.2712123139549676, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.610017776489258, "learning_rate": 9.033488766426451e-07, "loss": 0.4589, "mean_token_accuracy": 0.8502256870269775, "num_tokens": 81531421.0, "step": 2132 }, { "epoch": 0.27133952423355806, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.57324504852295, "learning_rate": 9.037727850784231e-07, "loss": 0.4249, "mean_token_accuracy": 0.8601003289222717, "num_tokens": 81567221.0, "step": 2133 }, { "epoch": 0.2714667345121486, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.612146377563477, "learning_rate": 9.041966935142009e-07, "loss": 0.4592, "mean_token_accuracy": 0.851844072341919, "num_tokens": 81605832.0, "step": 2134 }, { "epoch": 0.2715939447907391, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.537069320678711, "learning_rate": 9.046206019499788e-07, "loss": 0.4143, "mean_token_accuracy": 0.8670583963394165, "num_tokens": 81646919.0, "step": 2135 }, { "epoch": 0.2717211550693296, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.616960525512695, "learning_rate": 9.050445103857567e-07, "loss": 0.4923, "mean_token_accuracy": 0.8453797101974487, "num_tokens": 81687214.0, "step": 2136 }, { "epoch": 0.2718483653479201, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.578783988952637, "learning_rate": 9.054684188215344e-07, "loss": 0.3977, "mean_token_accuracy": 0.8713497519493103, "num_tokens": 81721437.0, "step": 2137 }, { "epoch": 0.27197557562651065, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.612217903137207, "learning_rate": 9.058923272573124e-07, "loss": 0.4717, "mean_token_accuracy": 0.8477174639701843, "num_tokens": 81756595.0, "step": 2138 }, { "epoch": 0.2721027859051011, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.574224472045898, "learning_rate": 9.063162356930902e-07, "loss": 0.4358, "mean_token_accuracy": 0.8598169088363647, "num_tokens": 81796198.0, "step": 2139 }, { "epoch": 0.27222999618369165, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.642383575439453, "learning_rate": 9.067401441288681e-07, "loss": 0.4673, "mean_token_accuracy": 0.8515309691429138, "num_tokens": 81833002.0, "step": 2140 }, { "epoch": 0.2723572064622822, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.624491691589355, "learning_rate": 9.07164052564646e-07, "loss": 0.4393, "mean_token_accuracy": 0.8532376885414124, "num_tokens": 81872089.0, "step": 2141 }, { "epoch": 0.27248441674087265, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.566804885864258, "learning_rate": 9.075879610004239e-07, "loss": 0.4746, "mean_token_accuracy": 0.8457592725753784, "num_tokens": 81908356.0, "step": 2142 }, { "epoch": 0.2726116270194632, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.606257438659668, "learning_rate": 9.080118694362017e-07, "loss": 0.4787, "mean_token_accuracy": 0.8432668447494507, "num_tokens": 81950251.0, "step": 2143 }, { "epoch": 0.2727388372980537, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.626879692077637, "learning_rate": 9.084357778719796e-07, "loss": 0.4713, "mean_token_accuracy": 0.8453050255775452, "num_tokens": 81987157.0, "step": 2144 }, { "epoch": 0.2728660475766442, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.652203559875488, "learning_rate": 9.088596863077574e-07, "loss": 0.4558, "mean_token_accuracy": 0.8559354543685913, "num_tokens": 82023633.0, "step": 2145 }, { "epoch": 0.2729932578552347, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.616011619567871, "learning_rate": 9.092835947435354e-07, "loss": 0.4578, "mean_token_accuracy": 0.8483962416648865, "num_tokens": 82061311.0, "step": 2146 }, { "epoch": 0.27312046813382523, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.583637237548828, "learning_rate": 9.097075031793132e-07, "loss": 0.4932, "mean_token_accuracy": 0.8411521315574646, "num_tokens": 82098370.0, "step": 2147 }, { "epoch": 0.2732476784124157, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.646796226501465, "learning_rate": 9.101314116150911e-07, "loss": 0.421, "mean_token_accuracy": 0.8674516677856445, "num_tokens": 82134931.0, "step": 2148 }, { "epoch": 0.27337488869100623, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.671182632446289, "learning_rate": 9.10555320050869e-07, "loss": 0.4301, "mean_token_accuracy": 0.8608532547950745, "num_tokens": 82171391.0, "step": 2149 }, { "epoch": 0.27350209896959676, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.581233024597168, "learning_rate": 9.109792284866469e-07, "loss": 0.4492, "mean_token_accuracy": 0.8582837581634521, "num_tokens": 82214064.0, "step": 2150 }, { "epoch": 0.27362930924818724, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.634149551391602, "learning_rate": 9.114031369224247e-07, "loss": 0.4645, "mean_token_accuracy": 0.8500503301620483, "num_tokens": 82252774.0, "step": 2151 }, { "epoch": 0.27375651952677776, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.661969184875488, "learning_rate": 9.118270453582026e-07, "loss": 0.52, "mean_token_accuracy": 0.8367693424224854, "num_tokens": 82291645.0, "step": 2152 }, { "epoch": 0.2738837298053683, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.640643119812012, "learning_rate": 9.122509537939804e-07, "loss": 0.452, "mean_token_accuracy": 0.8485358357429504, "num_tokens": 82333643.0, "step": 2153 }, { "epoch": 0.27401094008395877, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.683820724487305, "learning_rate": 9.126748622297584e-07, "loss": 0.4902, "mean_token_accuracy": 0.8380059599876404, "num_tokens": 82370792.0, "step": 2154 }, { "epoch": 0.2741381503625493, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.64651870727539, "learning_rate": 9.130987706655362e-07, "loss": 0.4797, "mean_token_accuracy": 0.8441365957260132, "num_tokens": 82408626.0, "step": 2155 }, { "epoch": 0.2742653606411398, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.635372161865234, "learning_rate": 9.135226791013141e-07, "loss": 0.4994, "mean_token_accuracy": 0.8445161581039429, "num_tokens": 82448401.0, "step": 2156 }, { "epoch": 0.2743925709197303, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.692780494689941, "learning_rate": 9.13946587537092e-07, "loss": 0.4566, "mean_token_accuracy": 0.847970724105835, "num_tokens": 82476644.0, "step": 2157 }, { "epoch": 0.2745197811983208, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.640387535095215, "learning_rate": 9.143704959728699e-07, "loss": 0.472, "mean_token_accuracy": 0.8427718877792358, "num_tokens": 82512722.0, "step": 2158 }, { "epoch": 0.27464699147691135, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.688218116760254, "learning_rate": 9.147944044086476e-07, "loss": 0.4919, "mean_token_accuracy": 0.8411220908164978, "num_tokens": 82545393.0, "step": 2159 }, { "epoch": 0.2747742017555018, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.680150985717773, "learning_rate": 9.152183128444255e-07, "loss": 0.4892, "mean_token_accuracy": 0.8472002148628235, "num_tokens": 82586329.0, "step": 2160 }, { "epoch": 0.27490141203409235, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.668973922729492, "learning_rate": 9.156422212802034e-07, "loss": 0.4962, "mean_token_accuracy": 0.8373619318008423, "num_tokens": 82625563.0, "step": 2161 }, { "epoch": 0.2750286223126829, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.67386245727539, "learning_rate": 9.160661297159813e-07, "loss": 0.4346, "mean_token_accuracy": 0.8587636351585388, "num_tokens": 82660030.0, "step": 2162 }, { "epoch": 0.27515583259127335, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.64447021484375, "learning_rate": 9.164900381517592e-07, "loss": 0.4563, "mean_token_accuracy": 0.8514748811721802, "num_tokens": 82698794.0, "step": 2163 }, { "epoch": 0.2752830428698639, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.652423858642578, "learning_rate": 9.16913946587537e-07, "loss": 0.4547, "mean_token_accuracy": 0.8499068021774292, "num_tokens": 82740015.0, "step": 2164 }, { "epoch": 0.2754102531484544, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.74551773071289, "learning_rate": 9.17337855023315e-07, "loss": 0.4677, "mean_token_accuracy": 0.8476733565330505, "num_tokens": 82775765.0, "step": 2165 }, { "epoch": 0.2755374634270449, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.661954879760742, "learning_rate": 9.177617634590928e-07, "loss": 0.4188, "mean_token_accuracy": 0.8626856803894043, "num_tokens": 82814324.0, "step": 2166 }, { "epoch": 0.2756646737056354, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.642415046691895, "learning_rate": 9.181856718948706e-07, "loss": 0.4065, "mean_token_accuracy": 0.8658404350280762, "num_tokens": 82851007.0, "step": 2167 }, { "epoch": 0.27579188398422594, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.701462745666504, "learning_rate": 9.186095803306485e-07, "loss": 0.4495, "mean_token_accuracy": 0.8532800078392029, "num_tokens": 82886846.0, "step": 2168 }, { "epoch": 0.2759190942628164, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.708934783935547, "learning_rate": 9.190334887664264e-07, "loss": 0.4876, "mean_token_accuracy": 0.8430041074752808, "num_tokens": 82929607.0, "step": 2169 }, { "epoch": 0.27604630454140694, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.690516471862793, "learning_rate": 9.194573972022043e-07, "loss": 0.545, "mean_token_accuracy": 0.8315294981002808, "num_tokens": 82965089.0, "step": 2170 }, { "epoch": 0.27617351481999747, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.723764419555664, "learning_rate": 9.198813056379822e-07, "loss": 0.5196, "mean_token_accuracy": 0.8321399092674255, "num_tokens": 83008345.0, "step": 2171 }, { "epoch": 0.27630072509858794, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.773896217346191, "learning_rate": 9.2030521407376e-07, "loss": 0.4745, "mean_token_accuracy": 0.843895673751831, "num_tokens": 83043763.0, "step": 2172 }, { "epoch": 0.27642793537717847, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.735260009765625, "learning_rate": 9.20729122509538e-07, "loss": 0.4176, "mean_token_accuracy": 0.8631721138954163, "num_tokens": 83081389.0, "step": 2173 }, { "epoch": 0.276555145655769, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.636974334716797, "learning_rate": 9.211530309453158e-07, "loss": 0.4572, "mean_token_accuracy": 0.849976658821106, "num_tokens": 83120547.0, "step": 2174 }, { "epoch": 0.27668235593435947, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.665044784545898, "learning_rate": 9.215769393810936e-07, "loss": 0.4625, "mean_token_accuracy": 0.851132869720459, "num_tokens": 83157229.0, "step": 2175 }, { "epoch": 0.27680956621295, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.742083549499512, "learning_rate": 9.220008478168715e-07, "loss": 0.543, "mean_token_accuracy": 0.8258421421051025, "num_tokens": 83198217.0, "step": 2176 }, { "epoch": 0.2769367764915405, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.683695793151855, "learning_rate": 9.224247562526494e-07, "loss": 0.4632, "mean_token_accuracy": 0.8482103943824768, "num_tokens": 83238750.0, "step": 2177 }, { "epoch": 0.277063986770131, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.739389419555664, "learning_rate": 9.228486646884273e-07, "loss": 0.4517, "mean_token_accuracy": 0.8524072170257568, "num_tokens": 83280138.0, "step": 2178 }, { "epoch": 0.2771911970487215, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.714345932006836, "learning_rate": 9.232725731242052e-07, "loss": 0.504, "mean_token_accuracy": 0.8355781435966492, "num_tokens": 83320633.0, "step": 2179 }, { "epoch": 0.27731840732731206, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.730792045593262, "learning_rate": 9.23696481559983e-07, "loss": 0.4875, "mean_token_accuracy": 0.8410519957542419, "num_tokens": 83358787.0, "step": 2180 }, { "epoch": 0.2774456176059026, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.646270751953125, "learning_rate": 9.24120389995761e-07, "loss": 0.4305, "mean_token_accuracy": 0.8599395751953125, "num_tokens": 83400609.0, "step": 2181 }, { "epoch": 0.27757282788449306, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.695097923278809, "learning_rate": 9.245442984315387e-07, "loss": 0.4471, "mean_token_accuracy": 0.8535316586494446, "num_tokens": 83436058.0, "step": 2182 }, { "epoch": 0.2777000381630836, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.713890075683594, "learning_rate": 9.249682068673165e-07, "loss": 0.4808, "mean_token_accuracy": 0.8420997858047485, "num_tokens": 83474542.0, "step": 2183 }, { "epoch": 0.2778272484416741, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.712699890136719, "learning_rate": 9.253921153030945e-07, "loss": 0.517, "mean_token_accuracy": 0.8328584432601929, "num_tokens": 83515551.0, "step": 2184 }, { "epoch": 0.2779544587202646, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.745255470275879, "learning_rate": 9.258160237388723e-07, "loss": 0.4324, "mean_token_accuracy": 0.8612285256385803, "num_tokens": 83551826.0, "step": 2185 }, { "epoch": 0.2780816689988551, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.749385833740234, "learning_rate": 9.262399321746503e-07, "loss": 0.4497, "mean_token_accuracy": 0.8532623648643494, "num_tokens": 83586598.0, "step": 2186 }, { "epoch": 0.27820887927744564, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.78383731842041, "learning_rate": 9.266638406104281e-07, "loss": 0.4996, "mean_token_accuracy": 0.8380742073059082, "num_tokens": 83625996.0, "step": 2187 }, { "epoch": 0.2783360895560361, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.70438003540039, "learning_rate": 9.27087749046206e-07, "loss": 0.4687, "mean_token_accuracy": 0.848680853843689, "num_tokens": 83657322.0, "step": 2188 }, { "epoch": 0.27846329983462664, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.774567604064941, "learning_rate": 9.275116574819839e-07, "loss": 0.4638, "mean_token_accuracy": 0.849391520023346, "num_tokens": 83693578.0, "step": 2189 }, { "epoch": 0.27859051011321717, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.771504402160645, "learning_rate": 9.279355659177617e-07, "loss": 0.4688, "mean_token_accuracy": 0.8495529890060425, "num_tokens": 83727672.0, "step": 2190 }, { "epoch": 0.27871772039180764, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.747016906738281, "learning_rate": 9.283594743535395e-07, "loss": 0.5123, "mean_token_accuracy": 0.8366315364837646, "num_tokens": 83761405.0, "step": 2191 }, { "epoch": 0.2788449306703982, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.770215034484863, "learning_rate": 9.287833827893175e-07, "loss": 0.4522, "mean_token_accuracy": 0.8490822911262512, "num_tokens": 83801564.0, "step": 2192 }, { "epoch": 0.2789721409489887, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.732805252075195, "learning_rate": 9.292072912250953e-07, "loss": 0.481, "mean_token_accuracy": 0.8430289626121521, "num_tokens": 83834260.0, "step": 2193 }, { "epoch": 0.2790993512275792, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.773651123046875, "learning_rate": 9.296311996608733e-07, "loss": 0.4094, "mean_token_accuracy": 0.8651419281959534, "num_tokens": 83873518.0, "step": 2194 }, { "epoch": 0.2792265615061697, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.748031616210938, "learning_rate": 9.300551080966511e-07, "loss": 0.5235, "mean_token_accuracy": 0.8323392868041992, "num_tokens": 83913902.0, "step": 2195 }, { "epoch": 0.27935377178476023, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.805893898010254, "learning_rate": 9.30479016532429e-07, "loss": 0.4493, "mean_token_accuracy": 0.8502950668334961, "num_tokens": 83947952.0, "step": 2196 }, { "epoch": 0.2794809820633507, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.707595825195312, "learning_rate": 9.309029249682068e-07, "loss": 0.4586, "mean_token_accuracy": 0.8516891002655029, "num_tokens": 83986057.0, "step": 2197 }, { "epoch": 0.27960819234194123, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.767781257629395, "learning_rate": 9.313268334039847e-07, "loss": 0.4689, "mean_token_accuracy": 0.8476964235305786, "num_tokens": 84023881.0, "step": 2198 }, { "epoch": 0.27973540262053176, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.715927124023438, "learning_rate": 9.317507418397625e-07, "loss": 0.4031, "mean_token_accuracy": 0.8694150447845459, "num_tokens": 84061480.0, "step": 2199 }, { "epoch": 0.27986261289912223, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.76147747039795, "learning_rate": 9.321746502755404e-07, "loss": 0.5049, "mean_token_accuracy": 0.8389379382133484, "num_tokens": 84101782.0, "step": 2200 }, { "epoch": 0.27998982317771276, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.81071949005127, "learning_rate": 9.325985587113183e-07, "loss": 0.464, "mean_token_accuracy": 0.8489730954170227, "num_tokens": 84136125.0, "step": 2201 }, { "epoch": 0.2801170334563033, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.84992790222168, "learning_rate": 9.330224671470962e-07, "loss": 0.4762, "mean_token_accuracy": 0.8494836091995239, "num_tokens": 84169748.0, "step": 2202 }, { "epoch": 0.28024424373489376, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.818933486938477, "learning_rate": 9.334463755828741e-07, "loss": 0.4908, "mean_token_accuracy": 0.8365362882614136, "num_tokens": 84201292.0, "step": 2203 }, { "epoch": 0.2803714540134843, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.805071830749512, "learning_rate": 9.338702840186519e-07, "loss": 0.4911, "mean_token_accuracy": 0.8384108543395996, "num_tokens": 84240652.0, "step": 2204 }, { "epoch": 0.2804986642920748, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.790096282958984, "learning_rate": 9.342941924544298e-07, "loss": 0.4824, "mean_token_accuracy": 0.8435448408126831, "num_tokens": 84280252.0, "step": 2205 }, { "epoch": 0.2806258745706653, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.78497314453125, "learning_rate": 9.347181008902076e-07, "loss": 0.4212, "mean_token_accuracy": 0.8656761050224304, "num_tokens": 84316505.0, "step": 2206 }, { "epoch": 0.2807530848492558, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.766621589660645, "learning_rate": 9.351420093259855e-07, "loss": 0.4241, "mean_token_accuracy": 0.8622633218765259, "num_tokens": 84349990.0, "step": 2207 }, { "epoch": 0.28088029512784635, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.819260597229004, "learning_rate": 9.355659177617634e-07, "loss": 0.4145, "mean_token_accuracy": 0.8626358509063721, "num_tokens": 84381066.0, "step": 2208 }, { "epoch": 0.2810075054064368, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.713096618652344, "learning_rate": 9.359898261975413e-07, "loss": 0.4064, "mean_token_accuracy": 0.8664804100990295, "num_tokens": 84422540.0, "step": 2209 }, { "epoch": 0.28113471568502735, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.857638359069824, "learning_rate": 9.364137346333192e-07, "loss": 0.5339, "mean_token_accuracy": 0.8294495344161987, "num_tokens": 84462158.0, "step": 2210 }, { "epoch": 0.2812619259636179, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.8067045211792, "learning_rate": 9.368376430690971e-07, "loss": 0.4234, "mean_token_accuracy": 0.8606560230255127, "num_tokens": 84495887.0, "step": 2211 }, { "epoch": 0.28138913624220835, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.828883171081543, "learning_rate": 9.372615515048749e-07, "loss": 0.4957, "mean_token_accuracy": 0.8411158323287964, "num_tokens": 84530176.0, "step": 2212 }, { "epoch": 0.2815163465207989, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.84105110168457, "learning_rate": 9.376854599406528e-07, "loss": 0.4971, "mean_token_accuracy": 0.8410670757293701, "num_tokens": 84574450.0, "step": 2213 }, { "epoch": 0.2816435567993894, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.817484855651855, "learning_rate": 9.381093683764306e-07, "loss": 0.4651, "mean_token_accuracy": 0.8494096994400024, "num_tokens": 84610661.0, "step": 2214 }, { "epoch": 0.2817707670779799, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.790374755859375, "learning_rate": 9.385332768122085e-07, "loss": 0.4464, "mean_token_accuracy": 0.8536452651023865, "num_tokens": 84651259.0, "step": 2215 }, { "epoch": 0.2818979773565704, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.800629615783691, "learning_rate": 9.389571852479864e-07, "loss": 0.4506, "mean_token_accuracy": 0.8543179035186768, "num_tokens": 84692389.0, "step": 2216 }, { "epoch": 0.28202518763516093, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.888838768005371, "learning_rate": 9.393810936837643e-07, "loss": 0.4545, "mean_token_accuracy": 0.8521960377693176, "num_tokens": 84731905.0, "step": 2217 }, { "epoch": 0.2821523979137514, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.777642250061035, "learning_rate": 9.398050021195422e-07, "loss": 0.4824, "mean_token_accuracy": 0.8480402231216431, "num_tokens": 84773727.0, "step": 2218 }, { "epoch": 0.28227960819234194, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.75187873840332, "learning_rate": 9.402289105553201e-07, "loss": 0.4282, "mean_token_accuracy": 0.8593699932098389, "num_tokens": 84816771.0, "step": 2219 }, { "epoch": 0.28240681847093246, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.883625030517578, "learning_rate": 9.406528189910978e-07, "loss": 0.4528, "mean_token_accuracy": 0.8533982038497925, "num_tokens": 84853172.0, "step": 2220 }, { "epoch": 0.28253402874952294, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.769432067871094, "learning_rate": 9.410767274268757e-07, "loss": 0.4328, "mean_token_accuracy": 0.8566566109657288, "num_tokens": 84891508.0, "step": 2221 }, { "epoch": 0.28266123902811346, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.81561279296875, "learning_rate": 9.415006358626536e-07, "loss": 0.5306, "mean_token_accuracy": 0.8279813528060913, "num_tokens": 84931204.0, "step": 2222 }, { "epoch": 0.282788449306704, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.904034614562988, "learning_rate": 9.419245442984314e-07, "loss": 0.5125, "mean_token_accuracy": 0.8350784778594971, "num_tokens": 84966782.0, "step": 2223 }, { "epoch": 0.28291565958529447, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.820334434509277, "learning_rate": 9.423484527342094e-07, "loss": 0.4579, "mean_token_accuracy": 0.8482072949409485, "num_tokens": 85001512.0, "step": 2224 }, { "epoch": 0.283042869863885, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.82996654510498, "learning_rate": 9.427723611699872e-07, "loss": 0.4662, "mean_token_accuracy": 0.8501758575439453, "num_tokens": 85036082.0, "step": 2225 }, { "epoch": 0.2831700801424755, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.840802192687988, "learning_rate": 9.431962696057652e-07, "loss": 0.4728, "mean_token_accuracy": 0.8473197817802429, "num_tokens": 85079433.0, "step": 2226 }, { "epoch": 0.283297290421066, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.822479248046875, "learning_rate": 9.43620178041543e-07, "loss": 0.424, "mean_token_accuracy": 0.8639850616455078, "num_tokens": 85116940.0, "step": 2227 }, { "epoch": 0.2834245006996565, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.846346855163574, "learning_rate": 9.440440864773208e-07, "loss": 0.4774, "mean_token_accuracy": 0.8477387428283691, "num_tokens": 85153328.0, "step": 2228 }, { "epoch": 0.28355171097824705, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.851625442504883, "learning_rate": 9.444679949130987e-07, "loss": 0.4788, "mean_token_accuracy": 0.8438370227813721, "num_tokens": 85194059.0, "step": 2229 }, { "epoch": 0.2836789212568376, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.87427806854248, "learning_rate": 9.448919033488766e-07, "loss": 0.4987, "mean_token_accuracy": 0.8410593271255493, "num_tokens": 85230820.0, "step": 2230 }, { "epoch": 0.28380613153542805, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.853005409240723, "learning_rate": 9.453158117846544e-07, "loss": 0.4806, "mean_token_accuracy": 0.8453301191329956, "num_tokens": 85268864.0, "step": 2231 }, { "epoch": 0.2839333418140186, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.87649917602539, "learning_rate": 9.457397202204324e-07, "loss": 0.507, "mean_token_accuracy": 0.8377133011817932, "num_tokens": 85311377.0, "step": 2232 }, { "epoch": 0.2840605520926091, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.846146583557129, "learning_rate": 9.461636286562102e-07, "loss": 0.4215, "mean_token_accuracy": 0.8648484945297241, "num_tokens": 85347082.0, "step": 2233 }, { "epoch": 0.2841877623711996, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.871670722961426, "learning_rate": 9.465875370919882e-07, "loss": 0.5096, "mean_token_accuracy": 0.8391112089157104, "num_tokens": 85384696.0, "step": 2234 }, { "epoch": 0.2843149726497901, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.945594787597656, "learning_rate": 9.470114455277659e-07, "loss": 0.4956, "mean_token_accuracy": 0.8395602703094482, "num_tokens": 85420192.0, "step": 2235 }, { "epoch": 0.28444218292838064, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.89687442779541, "learning_rate": 9.474353539635438e-07, "loss": 0.4461, "mean_token_accuracy": 0.8536073565483093, "num_tokens": 85462281.0, "step": 2236 }, { "epoch": 0.2845693932069711, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.863213539123535, "learning_rate": 9.478592623993217e-07, "loss": 0.4517, "mean_token_accuracy": 0.8521641492843628, "num_tokens": 85495191.0, "step": 2237 }, { "epoch": 0.28469660348556164, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.8637113571167, "learning_rate": 9.482831708350996e-07, "loss": 0.4495, "mean_token_accuracy": 0.8555333614349365, "num_tokens": 85533803.0, "step": 2238 }, { "epoch": 0.28482381376415217, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.888257026672363, "learning_rate": 9.487070792708775e-07, "loss": 0.4957, "mean_token_accuracy": 0.8426174521446228, "num_tokens": 85573424.0, "step": 2239 }, { "epoch": 0.28495102404274264, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.7982759475708, "learning_rate": 9.491309877066554e-07, "loss": 0.4795, "mean_token_accuracy": 0.845171332359314, "num_tokens": 85616423.0, "step": 2240 }, { "epoch": 0.28507823432133317, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.90351390838623, "learning_rate": 9.495548961424332e-07, "loss": 0.5049, "mean_token_accuracy": 0.8375441431999207, "num_tokens": 85658410.0, "step": 2241 }, { "epoch": 0.2852054445999237, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.867517471313477, "learning_rate": 9.499788045782111e-07, "loss": 0.4633, "mean_token_accuracy": 0.8489311933517456, "num_tokens": 85695767.0, "step": 2242 }, { "epoch": 0.28533265487851417, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.875121116638184, "learning_rate": 9.504027130139889e-07, "loss": 0.479, "mean_token_accuracy": 0.8447133898735046, "num_tokens": 85737771.0, "step": 2243 }, { "epoch": 0.2854598651571047, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.857386589050293, "learning_rate": 9.508266214497667e-07, "loss": 0.4262, "mean_token_accuracy": 0.8639860153198242, "num_tokens": 85774789.0, "step": 2244 }, { "epoch": 0.2855870754356952, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.844023704528809, "learning_rate": 9.512505298855447e-07, "loss": 0.4649, "mean_token_accuracy": 0.8516271710395813, "num_tokens": 85817833.0, "step": 2245 }, { "epoch": 0.2857142857142857, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.908504486083984, "learning_rate": 9.516744383213225e-07, "loss": 0.4093, "mean_token_accuracy": 0.8646597266197205, "num_tokens": 85854488.0, "step": 2246 }, { "epoch": 0.2858414959928762, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.892891883850098, "learning_rate": 9.520983467571005e-07, "loss": 0.4874, "mean_token_accuracy": 0.8438196182250977, "num_tokens": 85893746.0, "step": 2247 }, { "epoch": 0.28596870627146675, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.926511764526367, "learning_rate": 9.525222551928783e-07, "loss": 0.526, "mean_token_accuracy": 0.8363727927207947, "num_tokens": 85932798.0, "step": 2248 }, { "epoch": 0.2860959165500572, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.916487693786621, "learning_rate": 9.529461636286562e-07, "loss": 0.44, "mean_token_accuracy": 0.8567726612091064, "num_tokens": 85964446.0, "step": 2249 }, { "epoch": 0.28622312682864776, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.914412498474121, "learning_rate": 9.533700720644341e-07, "loss": 0.4182, "mean_token_accuracy": 0.8649123907089233, "num_tokens": 86000587.0, "step": 2250 }, { "epoch": 0.2863503371072383, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.902026176452637, "learning_rate": 9.537939805002118e-07, "loss": 0.4702, "mean_token_accuracy": 0.848273515701294, "num_tokens": 86042891.0, "step": 2251 }, { "epoch": 0.28647754738582876, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.90099811553955, "learning_rate": 9.542178889359898e-07, "loss": 0.448, "mean_token_accuracy": 0.8544082641601562, "num_tokens": 86079973.0, "step": 2252 }, { "epoch": 0.2866047576644193, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.930793762207031, "learning_rate": 9.546417973717677e-07, "loss": 0.4707, "mean_token_accuracy": 0.8496844172477722, "num_tokens": 86119767.0, "step": 2253 }, { "epoch": 0.2867319679430098, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.889176368713379, "learning_rate": 9.550657058075455e-07, "loss": 0.491, "mean_token_accuracy": 0.8434858322143555, "num_tokens": 86163176.0, "step": 2254 }, { "epoch": 0.2868591782216003, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.922609329223633, "learning_rate": 9.554896142433234e-07, "loss": 0.4855, "mean_token_accuracy": 0.8425028324127197, "num_tokens": 86201523.0, "step": 2255 }, { "epoch": 0.2869863885001908, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.903669357299805, "learning_rate": 9.559135226791012e-07, "loss": 0.5046, "mean_token_accuracy": 0.8334158062934875, "num_tokens": 86240277.0, "step": 2256 }, { "epoch": 0.28711359877878134, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.877663612365723, "learning_rate": 9.563374311148793e-07, "loss": 0.4948, "mean_token_accuracy": 0.8411478400230408, "num_tokens": 86289714.0, "step": 2257 }, { "epoch": 0.2872408090573718, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.988418579101562, "learning_rate": 9.56761339550657e-07, "loss": 0.446, "mean_token_accuracy": 0.8566120862960815, "num_tokens": 86320786.0, "step": 2258 }, { "epoch": 0.28736801933596234, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.909859657287598, "learning_rate": 9.57185247986435e-07, "loss": 0.4353, "mean_token_accuracy": 0.8590035438537598, "num_tokens": 86358002.0, "step": 2259 }, { "epoch": 0.28749522961455287, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.902143478393555, "learning_rate": 9.576091564222128e-07, "loss": 0.4038, "mean_token_accuracy": 0.8662994503974915, "num_tokens": 86395334.0, "step": 2260 }, { "epoch": 0.28762243989314334, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.852080345153809, "learning_rate": 9.580330648579906e-07, "loss": 0.4813, "mean_token_accuracy": 0.8418565392494202, "num_tokens": 86432616.0, "step": 2261 }, { "epoch": 0.2877496501717339, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.960587501525879, "learning_rate": 9.584569732937685e-07, "loss": 0.476, "mean_token_accuracy": 0.8446839451789856, "num_tokens": 86472293.0, "step": 2262 }, { "epoch": 0.2878768604503244, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.902226448059082, "learning_rate": 9.588808817295463e-07, "loss": 0.4658, "mean_token_accuracy": 0.8452550768852234, "num_tokens": 86506304.0, "step": 2263 }, { "epoch": 0.2880040707289149, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.946829795837402, "learning_rate": 9.593047901653242e-07, "loss": 0.4641, "mean_token_accuracy": 0.8479689359664917, "num_tokens": 86540520.0, "step": 2264 }, { "epoch": 0.2881312810075054, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.910697937011719, "learning_rate": 9.597286986011022e-07, "loss": 0.4247, "mean_token_accuracy": 0.8610016703605652, "num_tokens": 86576956.0, "step": 2265 }, { "epoch": 0.28825849128609593, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.947927474975586, "learning_rate": 9.601526070368799e-07, "loss": 0.509, "mean_token_accuracy": 0.8365054130554199, "num_tokens": 86613799.0, "step": 2266 }, { "epoch": 0.2883857015646864, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.968696594238281, "learning_rate": 9.60576515472658e-07, "loss": 0.449, "mean_token_accuracy": 0.8543479442596436, "num_tokens": 86650338.0, "step": 2267 }, { "epoch": 0.28851291184327693, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.899371147155762, "learning_rate": 9.610004239084358e-07, "loss": 0.4615, "mean_token_accuracy": 0.8494043350219727, "num_tokens": 86691605.0, "step": 2268 }, { "epoch": 0.28864012212186746, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.926802635192871, "learning_rate": 9.614243323442136e-07, "loss": 0.5295, "mean_token_accuracy": 0.8284707069396973, "num_tokens": 86728767.0, "step": 2269 }, { "epoch": 0.28876733240045793, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.908792495727539, "learning_rate": 9.618482407799915e-07, "loss": 0.4415, "mean_token_accuracy": 0.854070782661438, "num_tokens": 86765240.0, "step": 2270 }, { "epoch": 0.28889454267904846, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.943242073059082, "learning_rate": 9.622721492157693e-07, "loss": 0.4096, "mean_token_accuracy": 0.8668394684791565, "num_tokens": 86804042.0, "step": 2271 }, { "epoch": 0.289021752957639, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.932443618774414, "learning_rate": 9.626960576515472e-07, "loss": 0.4596, "mean_token_accuracy": 0.8515538573265076, "num_tokens": 86835220.0, "step": 2272 }, { "epoch": 0.28914896323622946, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.939684867858887, "learning_rate": 9.63119966087325e-07, "loss": 0.446, "mean_token_accuracy": 0.8524339199066162, "num_tokens": 86873414.0, "step": 2273 }, { "epoch": 0.28927617351482, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.939854621887207, "learning_rate": 9.635438745231029e-07, "loss": 0.4365, "mean_token_accuracy": 0.8590856790542603, "num_tokens": 86914272.0, "step": 2274 }, { "epoch": 0.2894033837934105, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.939763069152832, "learning_rate": 9.63967782958881e-07, "loss": 0.4771, "mean_token_accuracy": 0.8431453108787537, "num_tokens": 86952425.0, "step": 2275 }, { "epoch": 0.289530594072001, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.951424598693848, "learning_rate": 9.643916913946588e-07, "loss": 0.4368, "mean_token_accuracy": 0.8578637838363647, "num_tokens": 86987842.0, "step": 2276 }, { "epoch": 0.2896578043505915, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.921494483947754, "learning_rate": 9.648155998304366e-07, "loss": 0.4099, "mean_token_accuracy": 0.8656693696975708, "num_tokens": 87026750.0, "step": 2277 }, { "epoch": 0.28978501462918205, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.95804500579834, "learning_rate": 9.652395082662145e-07, "loss": 0.5135, "mean_token_accuracy": 0.8330149054527283, "num_tokens": 87069685.0, "step": 2278 }, { "epoch": 0.2899122249077726, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.947114944458008, "learning_rate": 9.656634167019923e-07, "loss": 0.4689, "mean_token_accuracy": 0.8435612320899963, "num_tokens": 87107054.0, "step": 2279 }, { "epoch": 0.29003943518636305, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.948530197143555, "learning_rate": 9.660873251377701e-07, "loss": 0.4756, "mean_token_accuracy": 0.8454010486602783, "num_tokens": 87146254.0, "step": 2280 }, { "epoch": 0.2901666454649536, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.935074806213379, "learning_rate": 9.66511233573548e-07, "loss": 0.5004, "mean_token_accuracy": 0.8419290781021118, "num_tokens": 87182404.0, "step": 2281 }, { "epoch": 0.2902938557435441, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.988381385803223, "learning_rate": 9.669351420093258e-07, "loss": 0.4323, "mean_token_accuracy": 0.8596784472465515, "num_tokens": 87216551.0, "step": 2282 }, { "epoch": 0.2904210660221346, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.972319602966309, "learning_rate": 9.67359050445104e-07, "loss": 0.4558, "mean_token_accuracy": 0.8501274585723877, "num_tokens": 87249402.0, "step": 2283 }, { "epoch": 0.2905482763007251, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.968245506286621, "learning_rate": 9.677829588808817e-07, "loss": 0.443, "mean_token_accuracy": 0.8532199263572693, "num_tokens": 87281846.0, "step": 2284 }, { "epoch": 0.29067548657931563, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 9.022340774536133, "learning_rate": 9.682068673166596e-07, "loss": 0.5132, "mean_token_accuracy": 0.8392860889434814, "num_tokens": 87315765.0, "step": 2285 }, { "epoch": 0.2908026968579061, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.959141731262207, "learning_rate": 9.686307757524374e-07, "loss": 0.4561, "mean_token_accuracy": 0.8510982394218445, "num_tokens": 87353668.0, "step": 2286 }, { "epoch": 0.29092990713649663, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.945651054382324, "learning_rate": 9.690546841882153e-07, "loss": 0.3878, "mean_token_accuracy": 0.8728225231170654, "num_tokens": 87390747.0, "step": 2287 }, { "epoch": 0.29105711741508716, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.960689544677734, "learning_rate": 9.694785926239931e-07, "loss": 0.4687, "mean_token_accuracy": 0.85092693567276, "num_tokens": 87430686.0, "step": 2288 }, { "epoch": 0.29118432769367764, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 9.041050910949707, "learning_rate": 9.69902501059771e-07, "loss": 0.508, "mean_token_accuracy": 0.8361005783081055, "num_tokens": 87465290.0, "step": 2289 }, { "epoch": 0.29131153797226816, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.944186210632324, "learning_rate": 9.703264094955488e-07, "loss": 0.4774, "mean_token_accuracy": 0.8447655439376831, "num_tokens": 87504553.0, "step": 2290 }, { "epoch": 0.2914387482508587, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.972228050231934, "learning_rate": 9.707503179313269e-07, "loss": 0.4478, "mean_token_accuracy": 0.8560600280761719, "num_tokens": 87544896.0, "step": 2291 }, { "epoch": 0.29156595852944917, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.936842918395996, "learning_rate": 9.711742263671047e-07, "loss": 0.4476, "mean_token_accuracy": 0.8544578552246094, "num_tokens": 87582075.0, "step": 2292 }, { "epoch": 0.2916931688080397, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.966004371643066, "learning_rate": 9.715981348028826e-07, "loss": 0.4408, "mean_token_accuracy": 0.8561856746673584, "num_tokens": 87623117.0, "step": 2293 }, { "epoch": 0.2918203790866302, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.9821195602417, "learning_rate": 9.720220432386604e-07, "loss": 0.4549, "mean_token_accuracy": 0.8534889817237854, "num_tokens": 87656263.0, "step": 2294 }, { "epoch": 0.2919475893652207, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.986102104187012, "learning_rate": 9.724459516744383e-07, "loss": 0.4288, "mean_token_accuracy": 0.8607972860336304, "num_tokens": 87695852.0, "step": 2295 }, { "epoch": 0.2920747996438112, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.971015930175781, "learning_rate": 9.728698601102161e-07, "loss": 0.4269, "mean_token_accuracy": 0.8607169985771179, "num_tokens": 87734331.0, "step": 2296 }, { "epoch": 0.29220200992240175, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.960957527160645, "learning_rate": 9.73293768545994e-07, "loss": 0.4609, "mean_token_accuracy": 0.8470566272735596, "num_tokens": 87768447.0, "step": 2297 }, { "epoch": 0.2923292202009922, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 9.0220308303833, "learning_rate": 9.737176769817718e-07, "loss": 0.488, "mean_token_accuracy": 0.8470203876495361, "num_tokens": 87813213.0, "step": 2298 }, { "epoch": 0.29245643047958275, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.983378410339355, "learning_rate": 9.741415854175499e-07, "loss": 0.5236, "mean_token_accuracy": 0.835821807384491, "num_tokens": 87851336.0, "step": 2299 }, { "epoch": 0.2925836407581733, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.989729881286621, "learning_rate": 9.745654938533277e-07, "loss": 0.4773, "mean_token_accuracy": 0.8450340628623962, "num_tokens": 87888474.0, "step": 2300 }, { "epoch": 0.29271085103676375, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 9.002448081970215, "learning_rate": 9.749894022891056e-07, "loss": 0.5023, "mean_token_accuracy": 0.8394666910171509, "num_tokens": 87928232.0, "step": 2301 }, { "epoch": 0.2928380613153543, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.059718132019043, "learning_rate": 9.754133107248834e-07, "loss": 0.4201, "mean_token_accuracy": 0.8618701696395874, "num_tokens": 87966982.0, "step": 2302 }, { "epoch": 0.2929652715939448, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.942774772644043, "learning_rate": 9.758372191606612e-07, "loss": 0.5061, "mean_token_accuracy": 0.8351474404335022, "num_tokens": 88003550.0, "step": 2303 }, { "epoch": 0.2930924818725353, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.995526313781738, "learning_rate": 9.76261127596439e-07, "loss": 0.4175, "mean_token_accuracy": 0.8641897439956665, "num_tokens": 88038038.0, "step": 2304 }, { "epoch": 0.2932196921511258, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 9.006519317626953, "learning_rate": 9.76685036032217e-07, "loss": 0.4887, "mean_token_accuracy": 0.8386391997337341, "num_tokens": 88074627.0, "step": 2305 }, { "epoch": 0.29334690242971634, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.987462043762207, "learning_rate": 9.771089444679948e-07, "loss": 0.4321, "mean_token_accuracy": 0.8589666485786438, "num_tokens": 88116772.0, "step": 2306 }, { "epoch": 0.2934741127083068, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 9.03795337677002, "learning_rate": 9.775328529037728e-07, "loss": 0.477, "mean_token_accuracy": 0.8455955982208252, "num_tokens": 88152888.0, "step": 2307 }, { "epoch": 0.29360132298689734, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 9.016968727111816, "learning_rate": 9.779567613395507e-07, "loss": 0.5064, "mean_token_accuracy": 0.8390316963195801, "num_tokens": 88184544.0, "step": 2308 }, { "epoch": 0.29372853326548787, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.9721040725708, "learning_rate": 9.783806697753285e-07, "loss": 0.4499, "mean_token_accuracy": 0.8552447557449341, "num_tokens": 88221603.0, "step": 2309 }, { "epoch": 0.29385574354407834, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.987802505493164, "learning_rate": 9.788045782111064e-07, "loss": 0.4249, "mean_token_accuracy": 0.8612463474273682, "num_tokens": 88265109.0, "step": 2310 }, { "epoch": 0.29398295382266887, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.043404579162598, "learning_rate": 9.792284866468842e-07, "loss": 0.4154, "mean_token_accuracy": 0.8619505167007446, "num_tokens": 88302756.0, "step": 2311 }, { "epoch": 0.2941101641012594, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.99070930480957, "learning_rate": 9.79652395082662e-07, "loss": 0.4144, "mean_token_accuracy": 0.8596540689468384, "num_tokens": 88335605.0, "step": 2312 }, { "epoch": 0.29423737437984987, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 9.049790382385254, "learning_rate": 9.8007630351844e-07, "loss": 0.4847, "mean_token_accuracy": 0.8444446325302124, "num_tokens": 88369447.0, "step": 2313 }, { "epoch": 0.2943645846584404, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.063688278198242, "learning_rate": 9.805002119542178e-07, "loss": 0.5147, "mean_token_accuracy": 0.833940863609314, "num_tokens": 88404442.0, "step": 2314 }, { "epoch": 0.2944917949370309, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.994929313659668, "learning_rate": 9.809241203899958e-07, "loss": 0.4604, "mean_token_accuracy": 0.8525956273078918, "num_tokens": 88445870.0, "step": 2315 }, { "epoch": 0.2946190052156214, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.072011947631836, "learning_rate": 9.813480288257737e-07, "loss": 0.4742, "mean_token_accuracy": 0.8417056798934937, "num_tokens": 88480994.0, "step": 2316 }, { "epoch": 0.2947462154942119, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.039796829223633, "learning_rate": 9.817719372615515e-07, "loss": 0.4305, "mean_token_accuracy": 0.8593926429748535, "num_tokens": 88519666.0, "step": 2317 }, { "epoch": 0.29487342577280246, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.980938911437988, "learning_rate": 9.821958456973294e-07, "loss": 0.4043, "mean_token_accuracy": 0.8661150336265564, "num_tokens": 88553855.0, "step": 2318 }, { "epoch": 0.29500063605139293, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.06090259552002, "learning_rate": 9.826197541331072e-07, "loss": 0.4203, "mean_token_accuracy": 0.864560604095459, "num_tokens": 88589053.0, "step": 2319 }, { "epoch": 0.29512784632998346, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.01870059967041, "learning_rate": 9.83043662568885e-07, "loss": 0.4786, "mean_token_accuracy": 0.845760703086853, "num_tokens": 88627067.0, "step": 2320 }, { "epoch": 0.295255056608574, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.992501258850098, "learning_rate": 9.83467571004663e-07, "loss": 0.4232, "mean_token_accuracy": 0.8616098761558533, "num_tokens": 88668030.0, "step": 2321 }, { "epoch": 0.29538226688716446, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.038562774658203, "learning_rate": 9.838914794404407e-07, "loss": 0.4214, "mean_token_accuracy": 0.8637486696243286, "num_tokens": 88708889.0, "step": 2322 }, { "epoch": 0.295509477165755, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.082993507385254, "learning_rate": 9.843153878762188e-07, "loss": 0.4889, "mean_token_accuracy": 0.8441575169563293, "num_tokens": 88745689.0, "step": 2323 }, { "epoch": 0.2956366874443455, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 9.011006355285645, "learning_rate": 9.847392963119966e-07, "loss": 0.4401, "mean_token_accuracy": 0.8573938012123108, "num_tokens": 88784667.0, "step": 2324 }, { "epoch": 0.295763897722936, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 9.046515464782715, "learning_rate": 9.851632047477745e-07, "loss": 0.453, "mean_token_accuracy": 0.8515840768814087, "num_tokens": 88827334.0, "step": 2325 }, { "epoch": 0.2958911080015265, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.9957275390625, "learning_rate": 9.855871131835523e-07, "loss": 0.3867, "mean_token_accuracy": 0.8718092441558838, "num_tokens": 88863744.0, "step": 2326 }, { "epoch": 0.29601831828011704, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.10818099975586, "learning_rate": 9.860110216193302e-07, "loss": 0.4188, "mean_token_accuracy": 0.8649111390113831, "num_tokens": 88903693.0, "step": 2327 }, { "epoch": 0.2961455285587075, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 9.057361602783203, "learning_rate": 9.86434930055108e-07, "loss": 0.4143, "mean_token_accuracy": 0.8638396859169006, "num_tokens": 88936668.0, "step": 2328 }, { "epoch": 0.29627273883729804, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.082061767578125, "learning_rate": 9.868588384908859e-07, "loss": 0.4237, "mean_token_accuracy": 0.8586958050727844, "num_tokens": 88970920.0, "step": 2329 }, { "epoch": 0.2963999491158886, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 9.037456512451172, "learning_rate": 9.872827469266637e-07, "loss": 0.4262, "mean_token_accuracy": 0.8580785393714905, "num_tokens": 89010856.0, "step": 2330 }, { "epoch": 0.2965271593944791, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2649765014648438e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.081300735473633, "learning_rate": 9.877066553624418e-07, "loss": 0.4352, "mean_token_accuracy": 0.857128918170929, "num_tokens": 89047109.0, "step": 2331 }, { "epoch": 0.2966543696730696, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.018356323242188, "learning_rate": 9.881305637982196e-07, "loss": 0.4435, "mean_token_accuracy": 0.8565613627433777, "num_tokens": 89087859.0, "step": 2332 }, { "epoch": 0.2967815799516601, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.155191421508789, "learning_rate": 9.885544722339975e-07, "loss": 0.4255, "mean_token_accuracy": 0.8598987460136414, "num_tokens": 89123601.0, "step": 2333 }, { "epoch": 0.29690879023025063, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.054641723632812, "learning_rate": 9.889783806697753e-07, "loss": 0.4091, "mean_token_accuracy": 0.8676033020019531, "num_tokens": 89160133.0, "step": 2334 }, { "epoch": 0.2970360005088411, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.073404312133789, "learning_rate": 9.894022891055532e-07, "loss": 0.4028, "mean_token_accuracy": 0.8688132762908936, "num_tokens": 89196731.0, "step": 2335 }, { "epoch": 0.29716321078743163, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.124275207519531, "learning_rate": 9.89826197541331e-07, "loss": 0.3918, "mean_token_accuracy": 0.8714879751205444, "num_tokens": 89230159.0, "step": 2336 }, { "epoch": 0.29729042106602216, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.09663200378418, "learning_rate": 9.902501059771089e-07, "loss": 0.4561, "mean_token_accuracy": 0.8501396179199219, "num_tokens": 89264648.0, "step": 2337 }, { "epoch": 0.29741763134461263, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.088229179382324, "learning_rate": 9.906740144128867e-07, "loss": 0.4549, "mean_token_accuracy": 0.849587619304657, "num_tokens": 89302505.0, "step": 2338 }, { "epoch": 0.29754484162320316, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.114041328430176, "learning_rate": 9.910979228486648e-07, "loss": 0.4138, "mean_token_accuracy": 0.8628702759742737, "num_tokens": 89338079.0, "step": 2339 }, { "epoch": 0.2976720519017937, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.103754997253418, "learning_rate": 9.915218312844426e-07, "loss": 0.4789, "mean_token_accuracy": 0.8496277332305908, "num_tokens": 89376717.0, "step": 2340 }, { "epoch": 0.29779926218038416, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.084587097167969, "learning_rate": 9.919457397202205e-07, "loss": 0.4691, "mean_token_accuracy": 0.8473106622695923, "num_tokens": 89416164.0, "step": 2341 }, { "epoch": 0.2979264724589747, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.15099048614502, "learning_rate": 9.923696481559983e-07, "loss": 0.4237, "mean_token_accuracy": 0.8623498678207397, "num_tokens": 89446491.0, "step": 2342 }, { "epoch": 0.2980536827375652, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.14985179901123, "learning_rate": 9.927935565917761e-07, "loss": 0.4709, "mean_token_accuracy": 0.844396710395813, "num_tokens": 89483637.0, "step": 2343 }, { "epoch": 0.2981808930161557, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.11089038848877, "learning_rate": 9.93217465027554e-07, "loss": 0.464, "mean_token_accuracy": 0.8532772064208984, "num_tokens": 89526532.0, "step": 2344 }, { "epoch": 0.2983081032947462, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.134737014770508, "learning_rate": 9.936413734633318e-07, "loss": 0.5254, "mean_token_accuracy": 0.8362737894058228, "num_tokens": 89558353.0, "step": 2345 }, { "epoch": 0.29843531357333675, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.104151725769043, "learning_rate": 9.940652818991097e-07, "loss": 0.4172, "mean_token_accuracy": 0.8634130358695984, "num_tokens": 89598727.0, "step": 2346 }, { "epoch": 0.2985625238519272, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.125224113464355, "learning_rate": 9.944891903348877e-07, "loss": 0.4845, "mean_token_accuracy": 0.8427600264549255, "num_tokens": 89640093.0, "step": 2347 }, { "epoch": 0.29868973413051775, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.111398696899414, "learning_rate": 9.949130987706656e-07, "loss": 0.4303, "mean_token_accuracy": 0.8592097759246826, "num_tokens": 89681401.0, "step": 2348 }, { "epoch": 0.2988169444091083, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.08872127532959, "learning_rate": 9.953370072064432e-07, "loss": 0.4227, "mean_token_accuracy": 0.8611433506011963, "num_tokens": 89720366.0, "step": 2349 }, { "epoch": 0.29894415468769875, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.10614013671875, "learning_rate": 9.957609156422213e-07, "loss": 0.4999, "mean_token_accuracy": 0.8388177156448364, "num_tokens": 89762234.0, "step": 2350 }, { "epoch": 0.2990713649662893, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.107392311096191, "learning_rate": 9.961848240779991e-07, "loss": 0.4471, "mean_token_accuracy": 0.8552657961845398, "num_tokens": 89792208.0, "step": 2351 }, { "epoch": 0.2991985752448798, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.141627311706543, "learning_rate": 9.96608732513777e-07, "loss": 0.4245, "mean_token_accuracy": 0.8655272126197815, "num_tokens": 89833246.0, "step": 2352 }, { "epoch": 0.2993257855234703, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.07840633392334, "learning_rate": 9.970326409495548e-07, "loss": 0.4827, "mean_token_accuracy": 0.842343807220459, "num_tokens": 89873271.0, "step": 2353 }, { "epoch": 0.2994529958020608, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.11169719696045, "learning_rate": 9.974565493853327e-07, "loss": 0.4975, "mean_token_accuracy": 0.8383219242095947, "num_tokens": 89913069.0, "step": 2354 }, { "epoch": 0.29958020608065133, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.15515422821045, "learning_rate": 9.978804578211107e-07, "loss": 0.4783, "mean_token_accuracy": 0.844192624092102, "num_tokens": 89948915.0, "step": 2355 }, { "epoch": 0.2997074163592418, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.1220121383667, "learning_rate": 9.983043662568886e-07, "loss": 0.4761, "mean_token_accuracy": 0.8474997282028198, "num_tokens": 89982123.0, "step": 2356 }, { "epoch": 0.29983462663783234, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.154747009277344, "learning_rate": 9.987282746926662e-07, "loss": 0.4272, "mean_token_accuracy": 0.8605754375457764, "num_tokens": 90018259.0, "step": 2357 }, { "epoch": 0.29996183691642286, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.10339069366455, "learning_rate": 9.991521831284443e-07, "loss": 0.4481, "mean_token_accuracy": 0.8550062775611877, "num_tokens": 90059041.0, "step": 2358 }, { "epoch": 0.30008904719501334, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.205671310424805, "learning_rate": 9.995760915642221e-07, "loss": 0.3972, "mean_token_accuracy": 0.870434582233429, "num_tokens": 90089463.0, "step": 2359 }, { "epoch": 0.30021625747360386, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.109447479248047, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8614627122879028, "num_tokens": 90125691.0, "step": 2360 }, { "epoch": 0.3003434677521944, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.138763427734375, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8509711027145386, "num_tokens": 90161778.0, "step": 2361 }, { "epoch": 0.30047067803078487, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.286384582519531, "learning_rate": 1e-06, "loss": 0.4997, "mean_token_accuracy": 0.8390723466873169, "num_tokens": 90202243.0, "step": 2362 }, { "epoch": 0.3005978883093754, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.076720237731934, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8451882004737854, "num_tokens": 90248924.0, "step": 2363 }, { "epoch": 0.3007250985879659, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.246942520141602, "learning_rate": 1e-06, "loss": 0.5646, "mean_token_accuracy": 0.8248105049133301, "num_tokens": 90281988.0, "step": 2364 }, { "epoch": 0.3008523088665564, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.219807624816895, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8471227884292603, "num_tokens": 90317050.0, "step": 2365 }, { "epoch": 0.3009795191451469, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.203021049499512, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8602333664894104, "num_tokens": 90358064.0, "step": 2366 }, { "epoch": 0.30110672942373745, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.151159286499023, "learning_rate": 1e-06, "loss": 0.4869, "mean_token_accuracy": 0.8416116237640381, "num_tokens": 90400692.0, "step": 2367 }, { "epoch": 0.3012339397023279, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.332207679748535, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8444339632987976, "num_tokens": 90432207.0, "step": 2368 }, { "epoch": 0.30136114998091845, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.155385971069336, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.850537896156311, "num_tokens": 90466081.0, "step": 2369 }, { "epoch": 0.301488360259509, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.242469787597656, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.859323263168335, "num_tokens": 90501963.0, "step": 2370 }, { "epoch": 0.30161557053809945, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.24196720123291, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.863863468170166, "num_tokens": 90540651.0, "step": 2371 }, { "epoch": 0.30174278081669, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.147969245910645, "learning_rate": 1e-06, "loss": 0.4793, "mean_token_accuracy": 0.8451266288757324, "num_tokens": 90585105.0, "step": 2372 }, { "epoch": 0.3018699910952805, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.224967956542969, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8498064875602722, "num_tokens": 90626773.0, "step": 2373 }, { "epoch": 0.301997201373871, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.130661964416504, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8665483593940735, "num_tokens": 90665974.0, "step": 2374 }, { "epoch": 0.3021244116524615, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.173989295959473, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8473988771438599, "num_tokens": 90706801.0, "step": 2375 }, { "epoch": 0.30225162193105204, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.197986602783203, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8482236266136169, "num_tokens": 90743635.0, "step": 2376 }, { "epoch": 0.3023788322096425, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.136186599731445, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8531941175460815, "num_tokens": 90782797.0, "step": 2377 }, { "epoch": 0.30250604248823304, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.189324378967285, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8452197313308716, "num_tokens": 90822661.0, "step": 2378 }, { "epoch": 0.30263325276682357, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.185147285461426, "learning_rate": 1e-06, "loss": 0.4975, "mean_token_accuracy": 0.8342883586883545, "num_tokens": 90857804.0, "step": 2379 }, { "epoch": 0.3027604630454141, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.22127914428711, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.850504457950592, "num_tokens": 90895778.0, "step": 2380 }, { "epoch": 0.30288767332400457, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.09576416015625, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.859169602394104, "num_tokens": 90934463.0, "step": 2381 }, { "epoch": 0.3030148836025951, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.171916961669922, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.857300877571106, "num_tokens": 90967888.0, "step": 2382 }, { "epoch": 0.3031420938811856, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.119343757629395, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8594470024108887, "num_tokens": 91005200.0, "step": 2383 }, { "epoch": 0.3032693041597761, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.216437339782715, "learning_rate": 1e-06, "loss": 0.5146, "mean_token_accuracy": 0.836146891117096, "num_tokens": 91037840.0, "step": 2384 }, { "epoch": 0.3033965144383666, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.184403419494629, "learning_rate": 1e-06, "loss": 0.4966, "mean_token_accuracy": 0.8392879366874695, "num_tokens": 91073336.0, "step": 2385 }, { "epoch": 0.30352372471695716, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.172530174255371, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8528901934623718, "num_tokens": 91107001.0, "step": 2386 }, { "epoch": 0.30365093499554763, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.24445915222168, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8599123954772949, "num_tokens": 91143937.0, "step": 2387 }, { "epoch": 0.30377814527413816, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.130763053894043, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.8437408208847046, "num_tokens": 91185391.0, "step": 2388 }, { "epoch": 0.3039053555527287, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.200986862182617, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8507037162780762, "num_tokens": 91227653.0, "step": 2389 }, { "epoch": 0.30403256583131916, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.19846248626709, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8723085522651672, "num_tokens": 91263610.0, "step": 2390 }, { "epoch": 0.3041597761099097, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.234697341918945, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8477265238761902, "num_tokens": 91296880.0, "step": 2391 }, { "epoch": 0.3042869863885002, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.18198013305664, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8402549028396606, "num_tokens": 91336242.0, "step": 2392 }, { "epoch": 0.3044141966670907, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.250550270080566, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8597673773765564, "num_tokens": 91376527.0, "step": 2393 }, { "epoch": 0.3045414069456812, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.216086387634277, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.843095064163208, "num_tokens": 91408415.0, "step": 2394 }, { "epoch": 0.30466861722427174, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.239969253540039, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8535164594650269, "num_tokens": 91449189.0, "step": 2395 }, { "epoch": 0.3047958275028622, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.209818840026855, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8485782146453857, "num_tokens": 91488599.0, "step": 2396 }, { "epoch": 0.30492303778145274, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.26519775390625, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8560712337493896, "num_tokens": 91525138.0, "step": 2397 }, { "epoch": 0.30505024806004327, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.143820762634277, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8544557690620422, "num_tokens": 91567444.0, "step": 2398 }, { "epoch": 0.30517745833863374, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.296009063720703, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.849042534828186, "num_tokens": 91602926.0, "step": 2399 }, { "epoch": 0.3053046686172243, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.205537796020508, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8637871742248535, "num_tokens": 91635901.0, "step": 2400 }, { "epoch": 0.3054318788958148, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.154346466064453, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8540438413619995, "num_tokens": 91679542.0, "step": 2401 }, { "epoch": 0.3055590891744053, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.192182540893555, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8475476503372192, "num_tokens": 91723022.0, "step": 2402 }, { "epoch": 0.3056862994529958, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.194514274597168, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8485093712806702, "num_tokens": 91760664.0, "step": 2403 }, { "epoch": 0.30581350973158633, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.17718505859375, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8480855226516724, "num_tokens": 91805570.0, "step": 2404 }, { "epoch": 0.3059407200101768, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.202704429626465, "learning_rate": 1e-06, "loss": 0.5009, "mean_token_accuracy": 0.8447898030281067, "num_tokens": 91843342.0, "step": 2405 }, { "epoch": 0.30606793028876733, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.233902931213379, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8431704640388489, "num_tokens": 91883872.0, "step": 2406 }, { "epoch": 0.30619514056735786, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.162796974182129, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8519843816757202, "num_tokens": 91923174.0, "step": 2407 }, { "epoch": 0.30632235084594833, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.241238594055176, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8508186340332031, "num_tokens": 91963462.0, "step": 2408 }, { "epoch": 0.30644956112453886, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.15757942199707, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8556908369064331, "num_tokens": 92000907.0, "step": 2409 }, { "epoch": 0.3065767714031294, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.247197151184082, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.855021595954895, "num_tokens": 92036110.0, "step": 2410 }, { "epoch": 0.30670398168171986, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.251134872436523, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8484305739402771, "num_tokens": 92080647.0, "step": 2411 }, { "epoch": 0.3068311919603104, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.175040245056152, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8563718795776367, "num_tokens": 92115407.0, "step": 2412 }, { "epoch": 0.3069584022389009, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.271289825439453, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8527547717094421, "num_tokens": 92149660.0, "step": 2413 }, { "epoch": 0.3070856125174914, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.176872253417969, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8551775217056274, "num_tokens": 92191581.0, "step": 2414 }, { "epoch": 0.3072128227960819, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.210770606994629, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.847706139087677, "num_tokens": 92230435.0, "step": 2415 }, { "epoch": 0.30734003307467245, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.180901527404785, "learning_rate": 1e-06, "loss": 0.4977, "mean_token_accuracy": 0.8404302000999451, "num_tokens": 92269545.0, "step": 2416 }, { "epoch": 0.3074672433532629, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.108959197998047, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8602409362792969, "num_tokens": 92307447.0, "step": 2417 }, { "epoch": 0.30759445363185345, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.286540985107422, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8511478304862976, "num_tokens": 92349441.0, "step": 2418 }, { "epoch": 0.307721663910444, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.191362380981445, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8572633266448975, "num_tokens": 92380405.0, "step": 2419 }, { "epoch": 0.30784887418903445, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.184429168701172, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8505248427391052, "num_tokens": 92419596.0, "step": 2420 }, { "epoch": 0.307976084467625, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.238137245178223, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.8461039066314697, "num_tokens": 92454590.0, "step": 2421 }, { "epoch": 0.3081032947462155, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.15535831451416, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8480339050292969, "num_tokens": 92491640.0, "step": 2422 }, { "epoch": 0.308230505024806, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.122929573059082, "learning_rate": 1e-06, "loss": 0.492, "mean_token_accuracy": 0.8440972566604614, "num_tokens": 92528682.0, "step": 2423 }, { "epoch": 0.3083577153033965, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.21196460723877, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8590841293334961, "num_tokens": 92566368.0, "step": 2424 }, { "epoch": 0.30848492558198704, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.17585563659668, "learning_rate": 1e-06, "loss": 0.521, "mean_token_accuracy": 0.8306128978729248, "num_tokens": 92601380.0, "step": 2425 }, { "epoch": 0.3086121358605775, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.176362037658691, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8677306175231934, "num_tokens": 92639724.0, "step": 2426 }, { "epoch": 0.30873934613916804, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.162203788757324, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8584380149841309, "num_tokens": 92677450.0, "step": 2427 }, { "epoch": 0.30886655641775856, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.162022590637207, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8605740070343018, "num_tokens": 92714681.0, "step": 2428 }, { "epoch": 0.3089937666963491, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.188973426818848, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8489575982093811, "num_tokens": 92749925.0, "step": 2429 }, { "epoch": 0.30912097697493957, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.169585227966309, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8565354347229004, "num_tokens": 92791720.0, "step": 2430 }, { "epoch": 0.3092481872535301, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.227076530456543, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8441985249519348, "num_tokens": 92826179.0, "step": 2431 }, { "epoch": 0.3093753975321206, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.214393615722656, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.852107048034668, "num_tokens": 92872751.0, "step": 2432 }, { "epoch": 0.3095026078107111, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.220881462097168, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.847220778465271, "num_tokens": 92907470.0, "step": 2433 }, { "epoch": 0.3096298180893016, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.230084419250488, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8577751517295837, "num_tokens": 92943321.0, "step": 2434 }, { "epoch": 0.30975702836789215, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.170369148254395, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8538293838500977, "num_tokens": 92982136.0, "step": 2435 }, { "epoch": 0.3098842386464826, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.248380661010742, "learning_rate": 1e-06, "loss": 0.4982, "mean_token_accuracy": 0.8425672054290771, "num_tokens": 93015974.0, "step": 2436 }, { "epoch": 0.31001144892507315, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.14859676361084, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8557919859886169, "num_tokens": 93054010.0, "step": 2437 }, { "epoch": 0.3101386592036637, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.26240348815918, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8604313135147095, "num_tokens": 93090226.0, "step": 2438 }, { "epoch": 0.31026586948225415, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.161055564880371, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8490114808082581, "num_tokens": 93130242.0, "step": 2439 }, { "epoch": 0.3103930797608447, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.209266662597656, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.84783935546875, "num_tokens": 93170990.0, "step": 2440 }, { "epoch": 0.3105202900394352, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.184605598449707, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8445779085159302, "num_tokens": 93205478.0, "step": 2441 }, { "epoch": 0.3106475003180257, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.226839065551758, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.84466153383255, "num_tokens": 93240952.0, "step": 2442 }, { "epoch": 0.3107747105966162, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.187406539916992, "learning_rate": 1e-06, "loss": 0.4909, "mean_token_accuracy": 0.8422430157661438, "num_tokens": 93278882.0, "step": 2443 }, { "epoch": 0.31090192087520674, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.219574928283691, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8440966606140137, "num_tokens": 93324573.0, "step": 2444 }, { "epoch": 0.3110291311537972, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.215795516967773, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8662902116775513, "num_tokens": 93361049.0, "step": 2445 }, { "epoch": 0.31115634143238774, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.210749626159668, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8485913276672363, "num_tokens": 93401748.0, "step": 2446 }, { "epoch": 0.31128355171097827, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.212709426879883, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8631511330604553, "num_tokens": 93434857.0, "step": 2447 }, { "epoch": 0.31141076198956874, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.225066184997559, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8499573469161987, "num_tokens": 93474681.0, "step": 2448 }, { "epoch": 0.31153797226815927, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.199247360229492, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8501040935516357, "num_tokens": 93510627.0, "step": 2449 }, { "epoch": 0.3116651825467498, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.238423347473145, "learning_rate": 1e-06, "loss": 0.4781, "mean_token_accuracy": 0.8430310487747192, "num_tokens": 93551598.0, "step": 2450 }, { "epoch": 0.31179239282534027, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.215834617614746, "learning_rate": 1e-06, "loss": 0.4842, "mean_token_accuracy": 0.8450814485549927, "num_tokens": 93586622.0, "step": 2451 }, { "epoch": 0.3119196031039308, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.176085472106934, "learning_rate": 1e-06, "loss": 0.5065, "mean_token_accuracy": 0.8413790464401245, "num_tokens": 93626496.0, "step": 2452 }, { "epoch": 0.3120468133825213, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.223211288452148, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8533135652542114, "num_tokens": 93665641.0, "step": 2453 }, { "epoch": 0.3121740236611118, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.227652549743652, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8577550649642944, "num_tokens": 93702135.0, "step": 2454 }, { "epoch": 0.3123012339397023, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.190324783325195, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8587901592254639, "num_tokens": 93738226.0, "step": 2455 }, { "epoch": 0.31242844421829286, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.184550285339355, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8546550869941711, "num_tokens": 93780273.0, "step": 2456 }, { "epoch": 0.31255565449688333, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.183409690856934, "learning_rate": 1e-06, "loss": 0.4956, "mean_token_accuracy": 0.839150071144104, "num_tokens": 93821977.0, "step": 2457 }, { "epoch": 0.31268286477547386, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.219694137573242, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8606182336807251, "num_tokens": 93856084.0, "step": 2458 }, { "epoch": 0.3128100750540644, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.231274604797363, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8550243973731995, "num_tokens": 93893491.0, "step": 2459 }, { "epoch": 0.31293728533265486, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.177054405212402, "learning_rate": 1e-06, "loss": 0.4923, "mean_token_accuracy": 0.8397455811500549, "num_tokens": 93940240.0, "step": 2460 }, { "epoch": 0.3130644956112454, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.231451988220215, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8445151448249817, "num_tokens": 93974504.0, "step": 2461 }, { "epoch": 0.3131917058898359, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.211297035217285, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8532024621963501, "num_tokens": 94009208.0, "step": 2462 }, { "epoch": 0.3133189161684264, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.200020790100098, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8517788052558899, "num_tokens": 94049249.0, "step": 2463 }, { "epoch": 0.3134461264470169, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.245779991149902, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8535367250442505, "num_tokens": 94085553.0, "step": 2464 }, { "epoch": 0.31357333672560744, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.212532997131348, "learning_rate": 1e-06, "loss": 0.5133, "mean_token_accuracy": 0.83189857006073, "num_tokens": 94127531.0, "step": 2465 }, { "epoch": 0.3137005470041979, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.208171844482422, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8635848760604858, "num_tokens": 94171191.0, "step": 2466 }, { "epoch": 0.31382775728278844, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.2361421585083, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8564445972442627, "num_tokens": 94212090.0, "step": 2467 }, { "epoch": 0.313954967561379, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.231996536254883, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8697574138641357, "num_tokens": 94246835.0, "step": 2468 }, { "epoch": 0.31408217783996945, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.205169677734375, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8500972390174866, "num_tokens": 94284581.0, "step": 2469 }, { "epoch": 0.31420938811856, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.241680145263672, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8617149591445923, "num_tokens": 94323050.0, "step": 2470 }, { "epoch": 0.3143365983971505, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.202917098999023, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8468290567398071, "num_tokens": 94364702.0, "step": 2471 }, { "epoch": 0.314463808675741, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.190825462341309, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8582766056060791, "num_tokens": 94404735.0, "step": 2472 }, { "epoch": 0.3145910189543315, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.143609046936035, "learning_rate": 1e-06, "loss": 0.5002, "mean_token_accuracy": 0.8372021913528442, "num_tokens": 94444683.0, "step": 2473 }, { "epoch": 0.31471822923292203, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.268548965454102, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8542314767837524, "num_tokens": 94484392.0, "step": 2474 }, { "epoch": 0.3148454395115125, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.155658721923828, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8587649464607239, "num_tokens": 94524860.0, "step": 2475 }, { "epoch": 0.31497264979010303, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.247002601623535, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.848662257194519, "num_tokens": 94565076.0, "step": 2476 }, { "epoch": 0.31509986006869356, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.16858196258545, "learning_rate": 1e-06, "loss": 0.5309, "mean_token_accuracy": 0.8292022943496704, "num_tokens": 94606191.0, "step": 2477 }, { "epoch": 0.31522707034728403, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.238777160644531, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.846450924873352, "num_tokens": 94648021.0, "step": 2478 }, { "epoch": 0.31535428062587456, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.239840507507324, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8628619909286499, "num_tokens": 94686305.0, "step": 2479 }, { "epoch": 0.3154814909044651, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.160499572753906, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8499277830123901, "num_tokens": 94728606.0, "step": 2480 }, { "epoch": 0.3156087011830556, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.218902587890625, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8549230098724365, "num_tokens": 94770162.0, "step": 2481 }, { "epoch": 0.3157359114616461, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.150806427001953, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.858368992805481, "num_tokens": 94807735.0, "step": 2482 }, { "epoch": 0.3158631217402366, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.241538047790527, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8470563292503357, "num_tokens": 94840366.0, "step": 2483 }, { "epoch": 0.31599033201882715, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.219825744628906, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8492665886878967, "num_tokens": 94874293.0, "step": 2484 }, { "epoch": 0.3161175422974176, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.235311508178711, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8519753813743591, "num_tokens": 94909347.0, "step": 2485 }, { "epoch": 0.31624475257600815, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.18004322052002, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8619790077209473, "num_tokens": 94955120.0, "step": 2486 }, { "epoch": 0.3163719628545987, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.228259086608887, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8662451505661011, "num_tokens": 94992319.0, "step": 2487 }, { "epoch": 0.31649917313318915, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.196986198425293, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8617985248565674, "num_tokens": 95023464.0, "step": 2488 }, { "epoch": 0.3166263834117797, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.258842468261719, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8652524948120117, "num_tokens": 95055963.0, "step": 2489 }, { "epoch": 0.3167535936903702, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.223484992980957, "learning_rate": 1e-06, "loss": 0.5126, "mean_token_accuracy": 0.8341559171676636, "num_tokens": 95089628.0, "step": 2490 }, { "epoch": 0.3168808039689607, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.168732643127441, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8664019107818604, "num_tokens": 95128557.0, "step": 2491 }, { "epoch": 0.3170080142475512, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.238283157348633, "learning_rate": 1e-06, "loss": 0.4936, "mean_token_accuracy": 0.8403264284133911, "num_tokens": 95168324.0, "step": 2492 }, { "epoch": 0.31713522452614173, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.192015647888184, "learning_rate": 1e-06, "loss": 0.5269, "mean_token_accuracy": 0.8326641321182251, "num_tokens": 95208139.0, "step": 2493 }, { "epoch": 0.3172624348047322, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.184749603271484, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8465820550918579, "num_tokens": 95247337.0, "step": 2494 }, { "epoch": 0.31738964508332274, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.19221305847168, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8447991609573364, "num_tokens": 95289208.0, "step": 2495 }, { "epoch": 0.31751685536191326, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.181158065795898, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8502456545829773, "num_tokens": 95329717.0, "step": 2496 }, { "epoch": 0.31764406564050374, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.167832374572754, "learning_rate": 1e-06, "loss": 0.5065, "mean_token_accuracy": 0.8322681188583374, "num_tokens": 95371052.0, "step": 2497 }, { "epoch": 0.31777127591909426, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.191739082336426, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8720529079437256, "num_tokens": 95412486.0, "step": 2498 }, { "epoch": 0.3178984861976848, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.227752685546875, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8688212633132935, "num_tokens": 95446167.0, "step": 2499 }, { "epoch": 0.31802569647627527, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.194170951843262, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8469924926757812, "num_tokens": 95485763.0, "step": 2500 }, { "epoch": 0.3181529067548658, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.220602989196777, "learning_rate": 1e-06, "loss": 0.5201, "mean_token_accuracy": 0.8421077728271484, "num_tokens": 95526407.0, "step": 2501 }, { "epoch": 0.3182801170334563, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.249651908874512, "learning_rate": 1e-06, "loss": 0.5619, "mean_token_accuracy": 0.8199398517608643, "num_tokens": 95567479.0, "step": 2502 }, { "epoch": 0.3184073273120468, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.193142890930176, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8589912056922913, "num_tokens": 95606588.0, "step": 2503 }, { "epoch": 0.3185345375906373, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5033950805664062e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.180692672729492, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8565468788146973, "num_tokens": 95653293.0, "step": 2504 }, { "epoch": 0.31866174786922785, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.242154121398926, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8431730270385742, "num_tokens": 95692424.0, "step": 2505 }, { "epoch": 0.3187889581478183, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.213784217834473, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8612813949584961, "num_tokens": 95733679.0, "step": 2506 }, { "epoch": 0.31891616842640885, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.228873252868652, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.847169041633606, "num_tokens": 95768211.0, "step": 2507 }, { "epoch": 0.3190433787049994, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.214763641357422, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8503183722496033, "num_tokens": 95805974.0, "step": 2508 }, { "epoch": 0.31917058898358985, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.271201133728027, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.850072979927063, "num_tokens": 95845463.0, "step": 2509 }, { "epoch": 0.3192977992621804, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.200251579284668, "learning_rate": 1e-06, "loss": 0.5079, "mean_token_accuracy": 0.8468549251556396, "num_tokens": 95888148.0, "step": 2510 }, { "epoch": 0.3194250095407709, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.228287696838379, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8740246295928955, "num_tokens": 95925432.0, "step": 2511 }, { "epoch": 0.3195522198193614, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.226227760314941, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8581178188323975, "num_tokens": 95970695.0, "step": 2512 }, { "epoch": 0.3196794300979519, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.23519515991211, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8526897430419922, "num_tokens": 96006093.0, "step": 2513 }, { "epoch": 0.31980664037654244, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.271097183227539, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8525757789611816, "num_tokens": 96045945.0, "step": 2514 }, { "epoch": 0.3199338506551329, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.23718547821045, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8690217733383179, "num_tokens": 96076748.0, "step": 2515 }, { "epoch": 0.32006106093372344, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.24720573425293, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8542808890342712, "num_tokens": 96116964.0, "step": 2516 }, { "epoch": 0.32018827121231397, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.279325485229492, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8547906875610352, "num_tokens": 96152073.0, "step": 2517 }, { "epoch": 0.32031548149090444, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.20129108428955, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8511898517608643, "num_tokens": 96191608.0, "step": 2518 }, { "epoch": 0.32044269176949497, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.331649780273438, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8697229623794556, "num_tokens": 96228124.0, "step": 2519 }, { "epoch": 0.3205699020480855, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.192741394042969, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8639159202575684, "num_tokens": 96265870.0, "step": 2520 }, { "epoch": 0.32069711232667597, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.289287567138672, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.861831545829773, "num_tokens": 96309623.0, "step": 2521 }, { "epoch": 0.3208243226052665, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.22087287902832, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8458571434020996, "num_tokens": 96344036.0, "step": 2522 }, { "epoch": 0.320951532883857, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.304462432861328, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.864022970199585, "num_tokens": 96377627.0, "step": 2523 }, { "epoch": 0.3210787431624475, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.253971099853516, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8541496992111206, "num_tokens": 96413568.0, "step": 2524 }, { "epoch": 0.32120595344103803, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.261394500732422, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8501957058906555, "num_tokens": 96454682.0, "step": 2525 }, { "epoch": 0.32133316371962856, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.262930870056152, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8418129682540894, "num_tokens": 96490456.0, "step": 2526 }, { "epoch": 0.32146037399821903, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.304258346557617, "learning_rate": 1e-06, "loss": 0.501, "mean_token_accuracy": 0.8462488055229187, "num_tokens": 96526891.0, "step": 2527 }, { "epoch": 0.32158758427680956, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.188826560974121, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8471031785011292, "num_tokens": 96575226.0, "step": 2528 }, { "epoch": 0.3217147945554001, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.31190299987793, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8428313732147217, "num_tokens": 96606383.0, "step": 2529 }, { "epoch": 0.3218420048339906, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.294890403747559, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8551200032234192, "num_tokens": 96639826.0, "step": 2530 }, { "epoch": 0.3219692151125811, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.211652755737305, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8572982549667358, "num_tokens": 96679151.0, "step": 2531 }, { "epoch": 0.3220964253911716, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.280988693237305, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8595790863037109, "num_tokens": 96719365.0, "step": 2532 }, { "epoch": 0.32222363566976214, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.240209579467773, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8503273129463196, "num_tokens": 96758318.0, "step": 2533 }, { "epoch": 0.3223508459483526, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.250609397888184, "learning_rate": 1e-06, "loss": 0.4886, "mean_token_accuracy": 0.8481351137161255, "num_tokens": 96791159.0, "step": 2534 }, { "epoch": 0.32247805622694314, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.30241584777832, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8554666042327881, "num_tokens": 96825352.0, "step": 2535 }, { "epoch": 0.32260526650553367, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.30715274810791, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8507213592529297, "num_tokens": 96859700.0, "step": 2536 }, { "epoch": 0.32273247678412414, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.25601863861084, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.850871741771698, "num_tokens": 96897077.0, "step": 2537 }, { "epoch": 0.3228596870627147, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.253582000732422, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8474042415618896, "num_tokens": 96936936.0, "step": 2538 }, { "epoch": 0.3229868973413052, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.272592544555664, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8584654927253723, "num_tokens": 96975702.0, "step": 2539 }, { "epoch": 0.3231141076198957, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.267120361328125, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8521952033042908, "num_tokens": 97010585.0, "step": 2540 }, { "epoch": 0.3232413178984862, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.261853218078613, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8771660923957825, "num_tokens": 97047210.0, "step": 2541 }, { "epoch": 0.32336852817707673, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.263527870178223, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8472751975059509, "num_tokens": 97084835.0, "step": 2542 }, { "epoch": 0.3234957384556672, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.291179656982422, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8503239750862122, "num_tokens": 97125829.0, "step": 2543 }, { "epoch": 0.32362294873425773, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.218559265136719, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.846275806427002, "num_tokens": 97162874.0, "step": 2544 }, { "epoch": 0.32375015901284826, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.235029220581055, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8561536073684692, "num_tokens": 97197559.0, "step": 2545 }, { "epoch": 0.32387736929143873, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.317083358764648, "learning_rate": 1e-06, "loss": 0.494, "mean_token_accuracy": 0.8397527933120728, "num_tokens": 97227028.0, "step": 2546 }, { "epoch": 0.32400457957002926, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.471220970153809, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8484401702880859, "num_tokens": 97270209.0, "step": 2547 }, { "epoch": 0.3241317898486198, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.257287979125977, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8560166358947754, "num_tokens": 97309176.0, "step": 2548 }, { "epoch": 0.32425900012721026, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.26794719696045, "learning_rate": 1e-06, "loss": 0.5109, "mean_token_accuracy": 0.8339199423789978, "num_tokens": 97340464.0, "step": 2549 }, { "epoch": 0.3243862104058008, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.208730697631836, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.855487585067749, "num_tokens": 97378435.0, "step": 2550 }, { "epoch": 0.3245134206843913, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.232357025146484, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8482052683830261, "num_tokens": 97417886.0, "step": 2551 }, { "epoch": 0.3246406309629818, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.262792587280273, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8559700846672058, "num_tokens": 97450552.0, "step": 2552 }, { "epoch": 0.3247678412415723, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.244881629943848, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8595219254493713, "num_tokens": 97488586.0, "step": 2553 }, { "epoch": 0.32489505152016285, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.276561737060547, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8476434946060181, "num_tokens": 97528893.0, "step": 2554 }, { "epoch": 0.3250222617987533, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.228864669799805, "learning_rate": 1e-06, "loss": 0.5273, "mean_token_accuracy": 0.8289191126823425, "num_tokens": 97568654.0, "step": 2555 }, { "epoch": 0.32514947207734385, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.199907302856445, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8565732836723328, "num_tokens": 97607068.0, "step": 2556 }, { "epoch": 0.3252766823559344, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.259325981140137, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8742521405220032, "num_tokens": 97647505.0, "step": 2557 }, { "epoch": 0.32540389263452485, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.215015411376953, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8468002080917358, "num_tokens": 97690505.0, "step": 2558 }, { "epoch": 0.3255311029131154, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.294441223144531, "learning_rate": 1e-06, "loss": 0.51, "mean_token_accuracy": 0.8401798605918884, "num_tokens": 97723955.0, "step": 2559 }, { "epoch": 0.3256583131917059, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.220606803894043, "learning_rate": 1e-06, "loss": 0.4724, "mean_token_accuracy": 0.8471310138702393, "num_tokens": 97763822.0, "step": 2560 }, { "epoch": 0.3257855234702964, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.264605522155762, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8479045629501343, "num_tokens": 97803827.0, "step": 2561 }, { "epoch": 0.3259127337488869, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.251811027526855, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8515228033065796, "num_tokens": 97841056.0, "step": 2562 }, { "epoch": 0.32603994402747744, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.224708557128906, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8695411682128906, "num_tokens": 97881714.0, "step": 2563 }, { "epoch": 0.3261671543060679, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.246135711669922, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8498623967170715, "num_tokens": 97925797.0, "step": 2564 }, { "epoch": 0.32629436458465844, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.302262306213379, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8568156957626343, "num_tokens": 97963638.0, "step": 2565 }, { "epoch": 0.32642157486324896, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.24618911743164, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8461384177207947, "num_tokens": 98002043.0, "step": 2566 }, { "epoch": 0.32654878514183944, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.314841270446777, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8547910451889038, "num_tokens": 98041031.0, "step": 2567 }, { "epoch": 0.32667599542042997, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.293842315673828, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8596134185791016, "num_tokens": 98075510.0, "step": 2568 }, { "epoch": 0.3268032056990205, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.265884399414062, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8581523895263672, "num_tokens": 98114162.0, "step": 2569 }, { "epoch": 0.32693041597761097, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.303244590759277, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.847877562046051, "num_tokens": 98148722.0, "step": 2570 }, { "epoch": 0.3270576262562015, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.309494018554688, "learning_rate": 1e-06, "loss": 0.5274, "mean_token_accuracy": 0.83211350440979, "num_tokens": 98188255.0, "step": 2571 }, { "epoch": 0.327184836534792, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.256468772888184, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8629616498947144, "num_tokens": 98228597.0, "step": 2572 }, { "epoch": 0.3273120468133825, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.214008331298828, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8435843586921692, "num_tokens": 98273405.0, "step": 2573 }, { "epoch": 0.327439257091973, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.333953857421875, "learning_rate": 1e-06, "loss": 0.5169, "mean_token_accuracy": 0.8382220268249512, "num_tokens": 98315091.0, "step": 2574 }, { "epoch": 0.32756646737056355, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.249089241027832, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8531208038330078, "num_tokens": 98356967.0, "step": 2575 }, { "epoch": 0.327693677649154, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.307326316833496, "learning_rate": 1e-06, "loss": 0.487, "mean_token_accuracy": 0.8432087302207947, "num_tokens": 98394455.0, "step": 2576 }, { "epoch": 0.32782088792774455, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.325364112854004, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8610703349113464, "num_tokens": 98430455.0, "step": 2577 }, { "epoch": 0.3279480982063351, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.248228073120117, "learning_rate": 1e-06, "loss": 0.4975, "mean_token_accuracy": 0.8379510641098022, "num_tokens": 98472218.0, "step": 2578 }, { "epoch": 0.3280753084849256, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.3240385055542, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8594158887863159, "num_tokens": 98509540.0, "step": 2579 }, { "epoch": 0.3282025187635161, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.2708101272583, "learning_rate": 1e-06, "loss": 0.4721, "mean_token_accuracy": 0.8511094450950623, "num_tokens": 98544761.0, "step": 2580 }, { "epoch": 0.3283297290421066, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.297761917114258, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8653964996337891, "num_tokens": 98580522.0, "step": 2581 }, { "epoch": 0.32845693932069714, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.321296691894531, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8607180714607239, "num_tokens": 98612988.0, "step": 2582 }, { "epoch": 0.3285841495992876, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.308616638183594, "learning_rate": 1e-06, "loss": 0.4872, "mean_token_accuracy": 0.8408476114273071, "num_tokens": 98654213.0, "step": 2583 }, { "epoch": 0.32871135987787814, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.296062469482422, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.855517566204071, "num_tokens": 98697168.0, "step": 2584 }, { "epoch": 0.32883857015646867, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.292872428894043, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8577673435211182, "num_tokens": 98735664.0, "step": 2585 }, { "epoch": 0.32896578043505914, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.235787391662598, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.847488284111023, "num_tokens": 98777257.0, "step": 2586 }, { "epoch": 0.32909299071364967, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.340814590454102, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8572155833244324, "num_tokens": 98812443.0, "step": 2587 }, { "epoch": 0.3292202009922402, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.30078411102295, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8609885573387146, "num_tokens": 98848231.0, "step": 2588 }, { "epoch": 0.32934741127083067, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.222892761230469, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8557811379432678, "num_tokens": 98886487.0, "step": 2589 }, { "epoch": 0.3294746215494212, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.347772598266602, "learning_rate": 1e-06, "loss": 0.482, "mean_token_accuracy": 0.8429900407791138, "num_tokens": 98924750.0, "step": 2590 }, { "epoch": 0.3296018318280117, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.274608612060547, "learning_rate": 1e-06, "loss": 0.5383, "mean_token_accuracy": 0.8305681943893433, "num_tokens": 98964151.0, "step": 2591 }, { "epoch": 0.3297290421066022, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.252532005310059, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8640567064285278, "num_tokens": 99000277.0, "step": 2592 }, { "epoch": 0.3298562523851927, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.334846496582031, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8431325554847717, "num_tokens": 99042335.0, "step": 2593 }, { "epoch": 0.32998346266378326, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.269068717956543, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8513059616088867, "num_tokens": 99082964.0, "step": 2594 }, { "epoch": 0.33011067294237373, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.316986083984375, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8752621412277222, "num_tokens": 99123440.0, "step": 2595 }, { "epoch": 0.33023788322096426, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.287899017333984, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8510388135910034, "num_tokens": 99167047.0, "step": 2596 }, { "epoch": 0.3303650934995548, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.262555122375488, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8498910665512085, "num_tokens": 99202594.0, "step": 2597 }, { "epoch": 0.33049230377814526, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.261197090148926, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8588953614234924, "num_tokens": 99244732.0, "step": 2598 }, { "epoch": 0.3306195140567358, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.324472427368164, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8632441163063049, "num_tokens": 99287161.0, "step": 2599 }, { "epoch": 0.3307467243353263, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.31578540802002, "learning_rate": 1e-06, "loss": 0.4724, "mean_token_accuracy": 0.8450400233268738, "num_tokens": 99322650.0, "step": 2600 }, { "epoch": 0.3308739346139168, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.336787223815918, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8534858226776123, "num_tokens": 99358529.0, "step": 2601 }, { "epoch": 0.3310011448925073, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.300564765930176, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8615700006484985, "num_tokens": 99394010.0, "step": 2602 }, { "epoch": 0.33112835517109784, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.297364234924316, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8558968305587769, "num_tokens": 99434764.0, "step": 2603 }, { "epoch": 0.3312555654496883, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.266647338867188, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8550071716308594, "num_tokens": 99481322.0, "step": 2604 }, { "epoch": 0.33138277572827884, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.321918487548828, "learning_rate": 1e-06, "loss": 0.4996, "mean_token_accuracy": 0.84353107213974, "num_tokens": 99517719.0, "step": 2605 }, { "epoch": 0.3315099860068694, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.388360977172852, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8554519414901733, "num_tokens": 99553662.0, "step": 2606 }, { "epoch": 0.33163719628545985, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.3705472946167, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8525017499923706, "num_tokens": 99589195.0, "step": 2607 }, { "epoch": 0.3317644065640504, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.297117233276367, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.855475127696991, "num_tokens": 99626092.0, "step": 2608 }, { "epoch": 0.3318916168426409, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.357090950012207, "learning_rate": 1e-06, "loss": 0.5253, "mean_token_accuracy": 0.8280749917030334, "num_tokens": 99667087.0, "step": 2609 }, { "epoch": 0.3320188271212314, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.315057754516602, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8476645350456238, "num_tokens": 99705929.0, "step": 2610 }, { "epoch": 0.3321460373998219, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.35069465637207, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8502441644668579, "num_tokens": 99742044.0, "step": 2611 }, { "epoch": 0.33227324767841243, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.342755317687988, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8557661175727844, "num_tokens": 99772621.0, "step": 2612 }, { "epoch": 0.3324004579570029, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.312067985534668, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8583722710609436, "num_tokens": 99814405.0, "step": 2613 }, { "epoch": 0.33252766823559343, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.341425895690918, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8639250993728638, "num_tokens": 99852197.0, "step": 2614 }, { "epoch": 0.33265487851418396, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.339731216430664, "learning_rate": 1e-06, "loss": 0.4736, "mean_token_accuracy": 0.8479493260383606, "num_tokens": 99894254.0, "step": 2615 }, { "epoch": 0.33278208879277443, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.322881698608398, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8473507165908813, "num_tokens": 99936230.0, "step": 2616 }, { "epoch": 0.33290929907136496, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.332049369812012, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.854041576385498, "num_tokens": 99974964.0, "step": 2617 }, { "epoch": 0.3330365093499555, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.271561622619629, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8442989587783813, "num_tokens": 100018750.0, "step": 2618 }, { "epoch": 0.33316371962854596, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.333998680114746, "learning_rate": 1e-06, "loss": 0.4771, "mean_token_accuracy": 0.8479077816009521, "num_tokens": 100054064.0, "step": 2619 }, { "epoch": 0.3332909299071365, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.27078628540039, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.858394980430603, "num_tokens": 100092549.0, "step": 2620 }, { "epoch": 0.333418140185727, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.357136726379395, "learning_rate": 1e-06, "loss": 0.5353, "mean_token_accuracy": 0.8276137113571167, "num_tokens": 100123973.0, "step": 2621 }, { "epoch": 0.3335453504643175, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.305482864379883, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8522852659225464, "num_tokens": 100162142.0, "step": 2622 }, { "epoch": 0.333672560742908, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.293338775634766, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8476897478103638, "num_tokens": 100199352.0, "step": 2623 }, { "epoch": 0.33379977102149855, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.323272705078125, "learning_rate": 1e-06, "loss": 0.5048, "mean_token_accuracy": 0.8372857570648193, "num_tokens": 100237446.0, "step": 2624 }, { "epoch": 0.333926981300089, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.333666801452637, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8573170900344849, "num_tokens": 100272986.0, "step": 2625 }, { "epoch": 0.33405419157867955, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.328357696533203, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8439834713935852, "num_tokens": 100304998.0, "step": 2626 }, { "epoch": 0.3341814018572701, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.295284271240234, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8495703935623169, "num_tokens": 100338473.0, "step": 2627 }, { "epoch": 0.33430861213586055, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.333966255187988, "learning_rate": 1e-06, "loss": 0.5197, "mean_token_accuracy": 0.8362730741500854, "num_tokens": 100372644.0, "step": 2628 }, { "epoch": 0.3344358224144511, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.311104774475098, "learning_rate": 1e-06, "loss": 0.5209, "mean_token_accuracy": 0.8330065011978149, "num_tokens": 100413248.0, "step": 2629 }, { "epoch": 0.3345630326930416, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.353421211242676, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8472693562507629, "num_tokens": 100448022.0, "step": 2630 }, { "epoch": 0.33469024297163213, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.244485855102539, "learning_rate": 1e-06, "loss": 0.4974, "mean_token_accuracy": 0.8409003019332886, "num_tokens": 100493222.0, "step": 2631 }, { "epoch": 0.3348174532502226, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.267922401428223, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8579748868942261, "num_tokens": 100534736.0, "step": 2632 }, { "epoch": 0.33494466352881314, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.309989929199219, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8602556586265564, "num_tokens": 100576754.0, "step": 2633 }, { "epoch": 0.33507187380740366, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.294740676879883, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8449248671531677, "num_tokens": 100611407.0, "step": 2634 }, { "epoch": 0.33519908408599414, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.292320251464844, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8587241172790527, "num_tokens": 100643531.0, "step": 2635 }, { "epoch": 0.33532629436458466, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.325520515441895, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8535541296005249, "num_tokens": 100678612.0, "step": 2636 }, { "epoch": 0.3354535046431752, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.27198314666748, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8566830158233643, "num_tokens": 100718873.0, "step": 2637 }, { "epoch": 0.33558071492176567, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.259135246276855, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8508214950561523, "num_tokens": 100759883.0, "step": 2638 }, { "epoch": 0.3357079252003562, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.332260131835938, "learning_rate": 1e-06, "loss": 0.5156, "mean_token_accuracy": 0.8317469954490662, "num_tokens": 100800022.0, "step": 2639 }, { "epoch": 0.3358351354789467, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.357604026794434, "learning_rate": 1e-06, "loss": 0.5111, "mean_token_accuracy": 0.8374022245407104, "num_tokens": 100837299.0, "step": 2640 }, { "epoch": 0.3359623457575372, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.330256462097168, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8493194580078125, "num_tokens": 100868433.0, "step": 2641 }, { "epoch": 0.3360895560361277, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.361660957336426, "learning_rate": 1e-06, "loss": 0.4968, "mean_token_accuracy": 0.8373253345489502, "num_tokens": 100907768.0, "step": 2642 }, { "epoch": 0.33621676631471825, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.3458890914917, "learning_rate": 1e-06, "loss": 0.5296, "mean_token_accuracy": 0.8296369910240173, "num_tokens": 100942723.0, "step": 2643 }, { "epoch": 0.3363439765933087, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.301654815673828, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.857396125793457, "num_tokens": 100984997.0, "step": 2644 }, { "epoch": 0.33647118687189925, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.269043922424316, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8669447898864746, "num_tokens": 101026290.0, "step": 2645 }, { "epoch": 0.3365983971504898, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.324112892150879, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8455625772476196, "num_tokens": 101066766.0, "step": 2646 }, { "epoch": 0.33672560742908025, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.328434944152832, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.856238067150116, "num_tokens": 101100833.0, "step": 2647 }, { "epoch": 0.3368528177076708, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.324444770812988, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8624536395072937, "num_tokens": 101145552.0, "step": 2648 }, { "epoch": 0.3369800279862613, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.281000137329102, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8542087078094482, "num_tokens": 101186644.0, "step": 2649 }, { "epoch": 0.3371072382648518, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.329288482666016, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8494583368301392, "num_tokens": 101223356.0, "step": 2650 }, { "epoch": 0.3372344485434423, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.344624519348145, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8642468452453613, "num_tokens": 101260527.0, "step": 2651 }, { "epoch": 0.33736165882203284, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.288873672485352, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8516579866409302, "num_tokens": 101305403.0, "step": 2652 }, { "epoch": 0.3374888691006233, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.29233169555664, "learning_rate": 1e-06, "loss": 0.5129, "mean_token_accuracy": 0.8383944034576416, "num_tokens": 101345858.0, "step": 2653 }, { "epoch": 0.33761607937921384, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.341552734375, "learning_rate": 1e-06, "loss": 0.5248, "mean_token_accuracy": 0.8290703296661377, "num_tokens": 101380321.0, "step": 2654 }, { "epoch": 0.33774328965780437, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.282550811767578, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8604192733764648, "num_tokens": 101419120.0, "step": 2655 }, { "epoch": 0.33787049993639484, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.340680122375488, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8544298410415649, "num_tokens": 101453272.0, "step": 2656 }, { "epoch": 0.33799771021498537, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.406972885131836, "learning_rate": 1e-06, "loss": 0.4994, "mean_token_accuracy": 0.8416083455085754, "num_tokens": 101486879.0, "step": 2657 }, { "epoch": 0.3381249204935759, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.314770698547363, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8447536826133728, "num_tokens": 101523691.0, "step": 2658 }, { "epoch": 0.33825213077216637, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.351739883422852, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8535749912261963, "num_tokens": 101555279.0, "step": 2659 }, { "epoch": 0.3383793410507569, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.315038681030273, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.869515597820282, "num_tokens": 101591780.0, "step": 2660 }, { "epoch": 0.3385065513293474, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.30170726776123, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8681050539016724, "num_tokens": 101631041.0, "step": 2661 }, { "epoch": 0.3386337616079379, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.320185661315918, "learning_rate": 1e-06, "loss": 0.4898, "mean_token_accuracy": 0.8416179418563843, "num_tokens": 101666315.0, "step": 2662 }, { "epoch": 0.33876097188652843, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.334908485412598, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8651461601257324, "num_tokens": 101704046.0, "step": 2663 }, { "epoch": 0.33888818216511896, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.337510108947754, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8627572655677795, "num_tokens": 101743076.0, "step": 2664 }, { "epoch": 0.33901539244370943, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.294607162475586, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8655499815940857, "num_tokens": 101784006.0, "step": 2665 }, { "epoch": 0.33914260272229996, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.371988296508789, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8590174317359924, "num_tokens": 101821845.0, "step": 2666 }, { "epoch": 0.3392698130008905, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.300833702087402, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8577241897583008, "num_tokens": 101861121.0, "step": 2667 }, { "epoch": 0.33939702327948096, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.373019218444824, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8495534062385559, "num_tokens": 101894330.0, "step": 2668 }, { "epoch": 0.3395242335580715, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.398279190063477, "learning_rate": 1e-06, "loss": 0.5039, "mean_token_accuracy": 0.8388417959213257, "num_tokens": 101934835.0, "step": 2669 }, { "epoch": 0.339651443836662, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.308393478393555, "learning_rate": 1e-06, "loss": 0.5128, "mean_token_accuracy": 0.8354728817939758, "num_tokens": 101970943.0, "step": 2670 }, { "epoch": 0.3397786541152525, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.432785987854004, "learning_rate": 1e-06, "loss": 0.4894, "mean_token_accuracy": 0.8409494161605835, "num_tokens": 102001211.0, "step": 2671 }, { "epoch": 0.339905864393843, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.321577072143555, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8589208126068115, "num_tokens": 102039399.0, "step": 2672 }, { "epoch": 0.34003307467243354, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.37271785736084, "learning_rate": 1e-06, "loss": 0.4907, "mean_token_accuracy": 0.842025101184845, "num_tokens": 102070635.0, "step": 2673 }, { "epoch": 0.340160284951024, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.32768726348877, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8575586080551147, "num_tokens": 102104590.0, "step": 2674 }, { "epoch": 0.34028749522961454, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.322890281677246, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.858348548412323, "num_tokens": 102149267.0, "step": 2675 }, { "epoch": 0.3404147055082051, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.352346420288086, "learning_rate": 1e-06, "loss": 0.5242, "mean_token_accuracy": 0.8361383080482483, "num_tokens": 102188089.0, "step": 2676 }, { "epoch": 0.34054191578679555, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.347497940063477, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8592327833175659, "num_tokens": 102224335.0, "step": 2677 }, { "epoch": 0.3406691260653861, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.373396873474121, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8564133048057556, "num_tokens": 102263421.0, "step": 2678 }, { "epoch": 0.3407963363439766, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.335772514343262, "learning_rate": 1e-06, "loss": 0.519, "mean_token_accuracy": 0.8354209661483765, "num_tokens": 102301133.0, "step": 2679 }, { "epoch": 0.34092354662256713, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.370681762695312, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8568211197853088, "num_tokens": 102338516.0, "step": 2680 }, { "epoch": 0.3410507569011576, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.298471450805664, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.8431133031845093, "num_tokens": 102376010.0, "step": 2681 }, { "epoch": 0.34117796717974813, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.3235502243042, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8729707598686218, "num_tokens": 102417080.0, "step": 2682 }, { "epoch": 0.34130517745833866, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.319910049438477, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8549915552139282, "num_tokens": 102458042.0, "step": 2683 }, { "epoch": 0.34143238773692913, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.390254020690918, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.860889196395874, "num_tokens": 102497152.0, "step": 2684 }, { "epoch": 0.34155959801551966, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.396100997924805, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.8512043952941895, "num_tokens": 102531486.0, "step": 2685 }, { "epoch": 0.3416868082941102, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.367609024047852, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8515135049819946, "num_tokens": 102571552.0, "step": 2686 }, { "epoch": 0.34181401857270066, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.401679039001465, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8591619729995728, "num_tokens": 102607338.0, "step": 2687 }, { "epoch": 0.3419412288512912, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.314470291137695, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8530832529067993, "num_tokens": 102654445.0, "step": 2688 }, { "epoch": 0.3420684391298817, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.358864784240723, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8614987134933472, "num_tokens": 102695265.0, "step": 2689 }, { "epoch": 0.3421956494084722, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.346733093261719, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8575388789176941, "num_tokens": 102728621.0, "step": 2690 }, { "epoch": 0.3423228596870627, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.366703033447266, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8467531204223633, "num_tokens": 102766378.0, "step": 2691 }, { "epoch": 0.34245006996565325, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.393659591674805, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8520330190658569, "num_tokens": 102806331.0, "step": 2692 }, { "epoch": 0.3425772802442437, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.375570297241211, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.85557621717453, "num_tokens": 102845314.0, "step": 2693 }, { "epoch": 0.34270449052283425, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.346153259277344, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8517647385597229, "num_tokens": 102888614.0, "step": 2694 }, { "epoch": 0.3428317008014248, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.344425201416016, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8474459648132324, "num_tokens": 102925988.0, "step": 2695 }, { "epoch": 0.34295891108001525, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.404769897460938, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8609288930892944, "num_tokens": 102963150.0, "step": 2696 }, { "epoch": 0.3430861213586058, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.379179954528809, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8605714440345764, "num_tokens": 102999164.0, "step": 2697 }, { "epoch": 0.3432133316371963, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.372754096984863, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.861223578453064, "num_tokens": 103034337.0, "step": 2698 }, { "epoch": 0.3433405419157868, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.314915657043457, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8594870567321777, "num_tokens": 103075062.0, "step": 2699 }, { "epoch": 0.3434677521943773, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.451026916503906, "learning_rate": 1e-06, "loss": 0.4819, "mean_token_accuracy": 0.8479697704315186, "num_tokens": 103113209.0, "step": 2700 }, { "epoch": 0.34359496247296784, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.383410453796387, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8576555252075195, "num_tokens": 103153117.0, "step": 2701 }, { "epoch": 0.3437221727515583, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.433243751525879, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8519035577774048, "num_tokens": 103192234.0, "step": 2702 }, { "epoch": 0.34384938303014884, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.346359252929688, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8501663208007812, "num_tokens": 103224958.0, "step": 2703 }, { "epoch": 0.34397659330873936, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.433530807495117, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.855220377445221, "num_tokens": 103259679.0, "step": 2704 }, { "epoch": 0.34410380358732984, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.329239845275879, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8451025485992432, "num_tokens": 103299987.0, "step": 2705 }, { "epoch": 0.34423101386592037, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.332427978515625, "learning_rate": 1e-06, "loss": 0.4981, "mean_token_accuracy": 0.8390558958053589, "num_tokens": 103344457.0, "step": 2706 }, { "epoch": 0.3443582241445109, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.375259399414062, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8588815927505493, "num_tokens": 103385173.0, "step": 2707 }, { "epoch": 0.34448543442310137, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.376062393188477, "learning_rate": 1e-06, "loss": 0.504, "mean_token_accuracy": 0.8374592065811157, "num_tokens": 103425972.0, "step": 2708 }, { "epoch": 0.3446126447016919, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.387487411499023, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8494964838027954, "num_tokens": 103466819.0, "step": 2709 }, { "epoch": 0.3447398549802824, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.335769653320312, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8417980074882507, "num_tokens": 103509660.0, "step": 2710 }, { "epoch": 0.3448670652588729, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.358617782592773, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8471987843513489, "num_tokens": 103544557.0, "step": 2711 }, { "epoch": 0.3449942755374634, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.334203720092773, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8624417781829834, "num_tokens": 103585132.0, "step": 2712 }, { "epoch": 0.34512148581605395, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.421451568603516, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8618313670158386, "num_tokens": 103626353.0, "step": 2713 }, { "epoch": 0.3452486960946444, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.306221961975098, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8559053540229797, "num_tokens": 103659035.0, "step": 2714 }, { "epoch": 0.34537590637323495, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.377148628234863, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8613346815109253, "num_tokens": 103699497.0, "step": 2715 }, { "epoch": 0.3455031166518255, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.373148918151855, "learning_rate": 1e-06, "loss": 0.4801, "mean_token_accuracy": 0.8421598672866821, "num_tokens": 103733685.0, "step": 2716 }, { "epoch": 0.34563032693041595, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.383484840393066, "learning_rate": 1e-06, "loss": 0.5052, "mean_token_accuracy": 0.8369179964065552, "num_tokens": 103769399.0, "step": 2717 }, { "epoch": 0.3457575372090065, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.364133834838867, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8604580163955688, "num_tokens": 103809078.0, "step": 2718 }, { "epoch": 0.345884747487597, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.277130126953125, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8700429797172546, "num_tokens": 103846764.0, "step": 2719 }, { "epoch": 0.3460119577661875, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.397093772888184, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8625222444534302, "num_tokens": 103883089.0, "step": 2720 }, { "epoch": 0.346139168044778, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.358620643615723, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8708012104034424, "num_tokens": 103915775.0, "step": 2721 }, { "epoch": 0.34626637832336854, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.373029708862305, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8483567237854004, "num_tokens": 103953703.0, "step": 2722 }, { "epoch": 0.346393588601959, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.345233917236328, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8436735272407532, "num_tokens": 103990775.0, "step": 2723 }, { "epoch": 0.34652079888054954, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.331636428833008, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8574608564376831, "num_tokens": 104025927.0, "step": 2724 }, { "epoch": 0.34664800915914007, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.379547119140625, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8449630737304688, "num_tokens": 104065860.0, "step": 2725 }, { "epoch": 0.34677521943773054, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.36620807647705, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8558653593063354, "num_tokens": 104102117.0, "step": 2726 }, { "epoch": 0.34690242971632107, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.368431091308594, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8496586084365845, "num_tokens": 104140896.0, "step": 2727 }, { "epoch": 0.3470296399949116, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.373995780944824, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.852372944355011, "num_tokens": 104181355.0, "step": 2728 }, { "epoch": 0.3471568502735021, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.369808197021484, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8553378582000732, "num_tokens": 104218564.0, "step": 2729 }, { "epoch": 0.3472840605520926, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.330829620361328, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8579062223434448, "num_tokens": 104256553.0, "step": 2730 }, { "epoch": 0.3474112708306831, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.347676277160645, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8670371174812317, "num_tokens": 104291150.0, "step": 2731 }, { "epoch": 0.34753848110927366, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.36507511138916, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8457953333854675, "num_tokens": 104327977.0, "step": 2732 }, { "epoch": 0.34766569138786413, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.387067794799805, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8504045009613037, "num_tokens": 104365167.0, "step": 2733 }, { "epoch": 0.34779290166645466, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.425707817077637, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.859521746635437, "num_tokens": 104400304.0, "step": 2734 }, { "epoch": 0.3479201119450452, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.430460929870605, "learning_rate": 1e-06, "loss": 0.5022, "mean_token_accuracy": 0.837873101234436, "num_tokens": 104437974.0, "step": 2735 }, { "epoch": 0.34804732222363566, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.41939640045166, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8574661612510681, "num_tokens": 104474474.0, "step": 2736 }, { "epoch": 0.3481745325022262, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.387847900390625, "learning_rate": 1e-06, "loss": 0.4997, "mean_token_accuracy": 0.839701771736145, "num_tokens": 104511004.0, "step": 2737 }, { "epoch": 0.3483017427808167, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.38896656036377, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8476599454879761, "num_tokens": 104551889.0, "step": 2738 }, { "epoch": 0.3484289530594072, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.428699493408203, "learning_rate": 1e-06, "loss": 0.5003, "mean_token_accuracy": 0.837043285369873, "num_tokens": 104591962.0, "step": 2739 }, { "epoch": 0.3485561633379977, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.432184219360352, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.851402997970581, "num_tokens": 104636201.0, "step": 2740 }, { "epoch": 0.34868337361658824, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.306001663208008, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8528850078582764, "num_tokens": 104673448.0, "step": 2741 }, { "epoch": 0.3488105838951787, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.433279037475586, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8614289164543152, "num_tokens": 104708879.0, "step": 2742 }, { "epoch": 0.34893779417376924, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.37924575805664, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8641352653503418, "num_tokens": 104747073.0, "step": 2743 }, { "epoch": 0.3490650044523598, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.315852165222168, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8531324863433838, "num_tokens": 104786950.0, "step": 2744 }, { "epoch": 0.34919221473095025, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.379765510559082, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8588363528251648, "num_tokens": 104828306.0, "step": 2745 }, { "epoch": 0.3493194250095408, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.395054817199707, "learning_rate": 1e-06, "loss": 0.4907, "mean_token_accuracy": 0.8463718891143799, "num_tokens": 104867030.0, "step": 2746 }, { "epoch": 0.3494466352881313, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.282332420349121, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8581641912460327, "num_tokens": 104909182.0, "step": 2747 }, { "epoch": 0.3495738455667218, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.430676460266113, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8495334982872009, "num_tokens": 104950648.0, "step": 2748 }, { "epoch": 0.3497010558453123, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.380873680114746, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8587572574615479, "num_tokens": 104994780.0, "step": 2749 }, { "epoch": 0.34982826612390283, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.386640548706055, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.861137866973877, "num_tokens": 105029469.0, "step": 2750 }, { "epoch": 0.3499554764024933, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.437270164489746, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8479735851287842, "num_tokens": 105066797.0, "step": 2751 }, { "epoch": 0.35008268668108383, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.314136505126953, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8692702054977417, "num_tokens": 105100869.0, "step": 2752 }, { "epoch": 0.35020989695967436, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.408873558044434, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8672479391098022, "num_tokens": 105135409.0, "step": 2753 }, { "epoch": 0.35033710723826483, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.4051513671875, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8611266613006592, "num_tokens": 105172839.0, "step": 2754 }, { "epoch": 0.35046431751685536, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.31140422821045, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8605027198791504, "num_tokens": 105216656.0, "step": 2755 }, { "epoch": 0.3505915277954459, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.393385887145996, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8491442203521729, "num_tokens": 105253740.0, "step": 2756 }, { "epoch": 0.35071873807403636, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.354649543762207, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8582921624183655, "num_tokens": 105290073.0, "step": 2757 }, { "epoch": 0.3508459483526269, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.397298812866211, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8611281514167786, "num_tokens": 105325933.0, "step": 2758 }, { "epoch": 0.3509731586312174, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.320183753967285, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8591905832290649, "num_tokens": 105362962.0, "step": 2759 }, { "epoch": 0.3511003689098079, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.329761505126953, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8552576303482056, "num_tokens": 105410074.0, "step": 2760 }, { "epoch": 0.3512275791883984, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.387019157409668, "learning_rate": 1e-06, "loss": 0.5132, "mean_token_accuracy": 0.8396206498146057, "num_tokens": 105450264.0, "step": 2761 }, { "epoch": 0.35135478946698895, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.338736534118652, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8460877537727356, "num_tokens": 105486566.0, "step": 2762 }, { "epoch": 0.3514819997455794, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.413790702819824, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8579602837562561, "num_tokens": 105527185.0, "step": 2763 }, { "epoch": 0.35160921002416995, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.392658233642578, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8531925678253174, "num_tokens": 105557401.0, "step": 2764 }, { "epoch": 0.3517364203027605, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.39819049835205, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8605008721351624, "num_tokens": 105586877.0, "step": 2765 }, { "epoch": 0.35186363058135095, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.371530532836914, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8592722415924072, "num_tokens": 105626988.0, "step": 2766 }, { "epoch": 0.3519908408599415, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.379206657409668, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8456094264984131, "num_tokens": 105665796.0, "step": 2767 }, { "epoch": 0.352118051138532, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.360845565795898, "learning_rate": 1e-06, "loss": 0.4749, "mean_token_accuracy": 0.8455187678337097, "num_tokens": 105708892.0, "step": 2768 }, { "epoch": 0.3522452614171225, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.34298038482666, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8485950827598572, "num_tokens": 105748221.0, "step": 2769 }, { "epoch": 0.352372471695713, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.506824493408203, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8523645401000977, "num_tokens": 105784568.0, "step": 2770 }, { "epoch": 0.35249968197430354, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.314360618591309, "learning_rate": 1e-06, "loss": 0.5028, "mean_token_accuracy": 0.8422012329101562, "num_tokens": 105827223.0, "step": 2771 }, { "epoch": 0.352626892252894, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.390331268310547, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8727066516876221, "num_tokens": 105871014.0, "step": 2772 }, { "epoch": 0.35275410253148454, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.407378196716309, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8507178425788879, "num_tokens": 105908995.0, "step": 2773 }, { "epoch": 0.35288131281007507, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.338818550109863, "learning_rate": 1e-06, "loss": 0.4842, "mean_token_accuracy": 0.8487275838851929, "num_tokens": 105948243.0, "step": 2774 }, { "epoch": 0.35300852308866554, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.423910140991211, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8564134836196899, "num_tokens": 105983997.0, "step": 2775 }, { "epoch": 0.35313573336725607, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.410420417785645, "learning_rate": 1e-06, "loss": 0.4846, "mean_token_accuracy": 0.841023862361908, "num_tokens": 106020336.0, "step": 2776 }, { "epoch": 0.3532629436458466, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.467825889587402, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.856101393699646, "num_tokens": 106053300.0, "step": 2777 }, { "epoch": 0.35339015392443707, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.289241790771484, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8684434294700623, "num_tokens": 106091790.0, "step": 2778 }, { "epoch": 0.3535173642030276, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.462981224060059, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8586089611053467, "num_tokens": 106123768.0, "step": 2779 }, { "epoch": 0.3536445744816181, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.407800674438477, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8628981113433838, "num_tokens": 106158219.0, "step": 2780 }, { "epoch": 0.35377178476020865, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.422027587890625, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8496620059013367, "num_tokens": 106195795.0, "step": 2781 }, { "epoch": 0.3538989950387991, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.431137084960938, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8551580309867859, "num_tokens": 106234973.0, "step": 2782 }, { "epoch": 0.35402620531738965, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.435013771057129, "learning_rate": 1e-06, "loss": 0.5108, "mean_token_accuracy": 0.8398633003234863, "num_tokens": 106267297.0, "step": 2783 }, { "epoch": 0.3541534155959802, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.36709976196289, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8618324398994446, "num_tokens": 106309070.0, "step": 2784 }, { "epoch": 0.35428062587457065, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.402215957641602, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8580869436264038, "num_tokens": 106345784.0, "step": 2785 }, { "epoch": 0.3544078361531612, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.406649589538574, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8621076345443726, "num_tokens": 106383172.0, "step": 2786 }, { "epoch": 0.3545350464317517, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.413981437683105, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8528253436088562, "num_tokens": 106416916.0, "step": 2787 }, { "epoch": 0.3546622567103422, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.47542953491211, "learning_rate": 1e-06, "loss": 0.5037, "mean_token_accuracy": 0.8392738103866577, "num_tokens": 106456969.0, "step": 2788 }, { "epoch": 0.3547894669889327, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.421122550964355, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8522424101829529, "num_tokens": 106490476.0, "step": 2789 }, { "epoch": 0.35491667726752324, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.464456558227539, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8554078340530396, "num_tokens": 106525990.0, "step": 2790 }, { "epoch": 0.3550438875461137, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.389138221740723, "learning_rate": 1e-06, "loss": 0.5492, "mean_token_accuracy": 0.8300988674163818, "num_tokens": 106562953.0, "step": 2791 }, { "epoch": 0.35517109782470424, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.45557975769043, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8574687242507935, "num_tokens": 106599142.0, "step": 2792 }, { "epoch": 0.35529830810329477, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.396313667297363, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8658902645111084, "num_tokens": 106636743.0, "step": 2793 }, { "epoch": 0.35542551838188524, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.396117210388184, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8594728708267212, "num_tokens": 106678992.0, "step": 2794 }, { "epoch": 0.35555272866047577, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.369033813476562, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8454586267471313, "num_tokens": 106719884.0, "step": 2795 }, { "epoch": 0.3556799389390663, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.430800437927246, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8505140542984009, "num_tokens": 106755224.0, "step": 2796 }, { "epoch": 0.35580714921765677, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.429977416992188, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8663440942764282, "num_tokens": 106789366.0, "step": 2797 }, { "epoch": 0.3559343594962473, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.44245433807373, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8449475169181824, "num_tokens": 106826382.0, "step": 2798 }, { "epoch": 0.3560615697748378, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.409241676330566, "learning_rate": 1e-06, "loss": 0.4835, "mean_token_accuracy": 0.8454816341400146, "num_tokens": 106860027.0, "step": 2799 }, { "epoch": 0.3561887800534283, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.449102401733398, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8621988296508789, "num_tokens": 106898684.0, "step": 2800 }, { "epoch": 0.35631599033201883, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.42425537109375, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8554257154464722, "num_tokens": 106934496.0, "step": 2801 }, { "epoch": 0.35644320061060936, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.410307884216309, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8682904243469238, "num_tokens": 106974335.0, "step": 2802 }, { "epoch": 0.35657041088919983, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.383561134338379, "learning_rate": 1e-06, "loss": 0.5533, "mean_token_accuracy": 0.8295392394065857, "num_tokens": 107013914.0, "step": 2803 }, { "epoch": 0.35669762116779036, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.405226707458496, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8575181365013123, "num_tokens": 107053747.0, "step": 2804 }, { "epoch": 0.3568248314463809, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.412468910217285, "learning_rate": 1e-06, "loss": 0.4832, "mean_token_accuracy": 0.8442114591598511, "num_tokens": 107095245.0, "step": 2805 }, { "epoch": 0.35695204172497136, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.418878555297852, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8548551797866821, "num_tokens": 107133627.0, "step": 2806 }, { "epoch": 0.3570792520035619, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.353497505187988, "learning_rate": 1e-06, "loss": 0.5194, "mean_token_accuracy": 0.8301456570625305, "num_tokens": 107170128.0, "step": 2807 }, { "epoch": 0.3572064622821524, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.383477210998535, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8644174337387085, "num_tokens": 107209137.0, "step": 2808 }, { "epoch": 0.3573336725607429, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.450221061706543, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8490476608276367, "num_tokens": 107244866.0, "step": 2809 }, { "epoch": 0.3574608828393334, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.387304306030273, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8538490533828735, "num_tokens": 107287399.0, "step": 2810 }, { "epoch": 0.35758809311792394, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.426877975463867, "learning_rate": 1e-06, "loss": 0.487, "mean_token_accuracy": 0.8443714380264282, "num_tokens": 107325627.0, "step": 2811 }, { "epoch": 0.3577153033965144, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.364298820495605, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8702642321586609, "num_tokens": 107359375.0, "step": 2812 }, { "epoch": 0.35784251367510495, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.392836570739746, "learning_rate": 1e-06, "loss": 0.4909, "mean_token_accuracy": 0.8430460691452026, "num_tokens": 107400286.0, "step": 2813 }, { "epoch": 0.3579697239536955, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.39852237701416, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.845410168170929, "num_tokens": 107431768.0, "step": 2814 }, { "epoch": 0.35809693423228595, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.377720832824707, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.863533079624176, "num_tokens": 107470875.0, "step": 2815 }, { "epoch": 0.3582241445108765, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.45775032043457, "learning_rate": 1e-06, "loss": 0.4808, "mean_token_accuracy": 0.8453847169876099, "num_tokens": 107510988.0, "step": 2816 }, { "epoch": 0.358351354789467, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.318095207214355, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8620945811271667, "num_tokens": 107552423.0, "step": 2817 }, { "epoch": 0.3584785650680575, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.382868766784668, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.846951961517334, "num_tokens": 107591693.0, "step": 2818 }, { "epoch": 0.358605775346648, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.41539478302002, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.858585774898529, "num_tokens": 107626017.0, "step": 2819 }, { "epoch": 0.35873298562523853, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.410350799560547, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8700977563858032, "num_tokens": 107664927.0, "step": 2820 }, { "epoch": 0.358860195903829, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.404032707214355, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8465638160705566, "num_tokens": 107702487.0, "step": 2821 }, { "epoch": 0.35898740618241953, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.417675971984863, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8673048615455627, "num_tokens": 107741263.0, "step": 2822 }, { "epoch": 0.35911461646101006, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.412372589111328, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.857438862323761, "num_tokens": 107781079.0, "step": 2823 }, { "epoch": 0.35924182673960053, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.39749526977539, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8726571798324585, "num_tokens": 107819309.0, "step": 2824 }, { "epoch": 0.35936903701819106, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.408127784729004, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8550612330436707, "num_tokens": 107853251.0, "step": 2825 }, { "epoch": 0.3594962472967816, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.446687698364258, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8741652965545654, "num_tokens": 107893476.0, "step": 2826 }, { "epoch": 0.35962345757537206, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.391885757446289, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8636404275894165, "num_tokens": 107931204.0, "step": 2827 }, { "epoch": 0.3597506678539626, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.48141860961914, "learning_rate": 1e-06, "loss": 0.5031, "mean_token_accuracy": 0.8380168080329895, "num_tokens": 107966471.0, "step": 2828 }, { "epoch": 0.3598778781325531, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.38329792022705, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8556398153305054, "num_tokens": 108005328.0, "step": 2829 }, { "epoch": 0.36000508841114365, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.439501762390137, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8439979553222656, "num_tokens": 108046295.0, "step": 2830 }, { "epoch": 0.3601322986897341, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.50676441192627, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8566299080848694, "num_tokens": 108078164.0, "step": 2831 }, { "epoch": 0.36025950896832465, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.442039489746094, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8498851656913757, "num_tokens": 108117105.0, "step": 2832 }, { "epoch": 0.3603867192469152, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.458626747131348, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.852371096611023, "num_tokens": 108155301.0, "step": 2833 }, { "epoch": 0.36051392952550565, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.423426628112793, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.865023136138916, "num_tokens": 108193091.0, "step": 2834 }, { "epoch": 0.3606411398040962, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.420472145080566, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8529114723205566, "num_tokens": 108233719.0, "step": 2835 }, { "epoch": 0.3607683500826867, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.493022918701172, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8598161935806274, "num_tokens": 108265775.0, "step": 2836 }, { "epoch": 0.3608955603612772, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.393804550170898, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8643383383750916, "num_tokens": 108307953.0, "step": 2837 }, { "epoch": 0.3610227706398677, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.381779670715332, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8501750826835632, "num_tokens": 108351998.0, "step": 2838 }, { "epoch": 0.36114998091845824, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.47006893157959, "learning_rate": 1e-06, "loss": 0.4938, "mean_token_accuracy": 0.8425766825675964, "num_tokens": 108395968.0, "step": 2839 }, { "epoch": 0.3612771911970487, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.450972557067871, "learning_rate": 1e-06, "loss": 0.517, "mean_token_accuracy": 0.8360885381698608, "num_tokens": 108433045.0, "step": 2840 }, { "epoch": 0.36140440147563924, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.49616813659668, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.848924994468689, "num_tokens": 108470964.0, "step": 2841 }, { "epoch": 0.36153161175422976, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.45704174041748, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8549551963806152, "num_tokens": 108512572.0, "step": 2842 }, { "epoch": 0.36165882203282024, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.50113582611084, "learning_rate": 1e-06, "loss": 0.4867, "mean_token_accuracy": 0.8402619361877441, "num_tokens": 108548526.0, "step": 2843 }, { "epoch": 0.36178603231141077, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.459834098815918, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8511315584182739, "num_tokens": 108588447.0, "step": 2844 }, { "epoch": 0.3619132425900013, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.509546279907227, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8447849750518799, "num_tokens": 108628426.0, "step": 2845 }, { "epoch": 0.36204045286859177, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.419382095336914, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8498305082321167, "num_tokens": 108661355.0, "step": 2846 }, { "epoch": 0.3621676631471823, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.446650505065918, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8707364201545715, "num_tokens": 108698505.0, "step": 2847 }, { "epoch": 0.3622948734257728, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.444070816040039, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8590712547302246, "num_tokens": 108737409.0, "step": 2848 }, { "epoch": 0.3624220837043633, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.455721855163574, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8549190163612366, "num_tokens": 108772969.0, "step": 2849 }, { "epoch": 0.3625492939829538, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.548316955566406, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8603524565696716, "num_tokens": 108805129.0, "step": 2850 }, { "epoch": 0.36267650426154435, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.439911842346191, "learning_rate": 1e-06, "loss": 0.4905, "mean_token_accuracy": 0.8399158716201782, "num_tokens": 108841197.0, "step": 2851 }, { "epoch": 0.3628037145401348, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.482550621032715, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.868293046951294, "num_tokens": 108876442.0, "step": 2852 }, { "epoch": 0.36293092481872535, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.44784927368164, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8400450348854065, "num_tokens": 108911432.0, "step": 2853 }, { "epoch": 0.3630581350973159, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.468461990356445, "learning_rate": 1e-06, "loss": 0.4973, "mean_token_accuracy": 0.8400462865829468, "num_tokens": 108952977.0, "step": 2854 }, { "epoch": 0.36318534537590635, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.427044868469238, "learning_rate": 1e-06, "loss": 0.5342, "mean_token_accuracy": 0.8312625288963318, "num_tokens": 108995602.0, "step": 2855 }, { "epoch": 0.3633125556544969, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.499551773071289, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.856123685836792, "num_tokens": 109040226.0, "step": 2856 }, { "epoch": 0.3634397659330874, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.457709312438965, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8563331365585327, "num_tokens": 109082898.0, "step": 2857 }, { "epoch": 0.3635669762116779, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.479408264160156, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8446931838989258, "num_tokens": 109122460.0, "step": 2858 }, { "epoch": 0.3636941864902684, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.588298797607422, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8532916307449341, "num_tokens": 109161680.0, "step": 2859 }, { "epoch": 0.36382139676885894, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.511631965637207, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8736920356750488, "num_tokens": 109198943.0, "step": 2860 }, { "epoch": 0.3639486070474494, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.457812309265137, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.862510085105896, "num_tokens": 109235719.0, "step": 2861 }, { "epoch": 0.36407581732603994, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.493926048278809, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8467977046966553, "num_tokens": 109272681.0, "step": 2862 }, { "epoch": 0.36420302760463047, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.512909889221191, "learning_rate": 1e-06, "loss": 0.4963, "mean_token_accuracy": 0.8430650234222412, "num_tokens": 109306114.0, "step": 2863 }, { "epoch": 0.36433023788322094, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.522838592529297, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8456388115882874, "num_tokens": 109337728.0, "step": 2864 }, { "epoch": 0.36445744816181147, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.460914611816406, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8447310924530029, "num_tokens": 109377809.0, "step": 2865 }, { "epoch": 0.364584658440402, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.395591735839844, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8685881495475769, "num_tokens": 109413499.0, "step": 2866 }, { "epoch": 0.36471186871899247, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.530163764953613, "learning_rate": 1e-06, "loss": 0.4804, "mean_token_accuracy": 0.8444970846176147, "num_tokens": 109460961.0, "step": 2867 }, { "epoch": 0.364839078997583, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.425436019897461, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.864270806312561, "num_tokens": 109500615.0, "step": 2868 }, { "epoch": 0.36496628927617353, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.483323097229004, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8572861552238464, "num_tokens": 109533921.0, "step": 2869 }, { "epoch": 0.365093499554764, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.546367645263672, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8446744084358215, "num_tokens": 109567069.0, "step": 2870 }, { "epoch": 0.36522070983335453, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.462475776672363, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8671178817749023, "num_tokens": 109603696.0, "step": 2871 }, { "epoch": 0.36534792011194506, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.42723560333252, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8714994192123413, "num_tokens": 109644590.0, "step": 2872 }, { "epoch": 0.36547513039053553, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.431684494018555, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8497377038002014, "num_tokens": 109678869.0, "step": 2873 }, { "epoch": 0.36560234066912606, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.494240760803223, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8484521508216858, "num_tokens": 109722517.0, "step": 2874 }, { "epoch": 0.3657295509477166, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.489151000976562, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8508504629135132, "num_tokens": 109757869.0, "step": 2875 }, { "epoch": 0.36585676122630706, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.462044715881348, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8494351506233215, "num_tokens": 109793162.0, "step": 2876 }, { "epoch": 0.3659839715048976, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.462850570678711, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8607689142227173, "num_tokens": 109832942.0, "step": 2877 }, { "epoch": 0.3661111817834881, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.448381423950195, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8529161214828491, "num_tokens": 109876230.0, "step": 2878 }, { "epoch": 0.3662383920620786, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.612981796264648, "learning_rate": 1e-06, "loss": 0.4854, "mean_token_accuracy": 0.8441680669784546, "num_tokens": 109911844.0, "step": 2879 }, { "epoch": 0.3663656023406691, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9921531677246094e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.496512413024902, "learning_rate": 1e-06, "loss": 0.5127, "mean_token_accuracy": 0.8424150347709656, "num_tokens": 109957619.0, "step": 2880 }, { "epoch": 0.36649281261925964, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9921531677246094e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.68784236907959, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8427035808563232, "num_tokens": 109999117.0, "step": 2881 }, { "epoch": 0.3666200228978502, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.463634490966797, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8612698912620544, "num_tokens": 110039557.0, "step": 2882 }, { "epoch": 0.36674723317644065, "ewc_loss": 0.01416015625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.38835334777832, "learning_rate": 1e-06, "loss": 0.4721, "mean_token_accuracy": 0.8464067578315735, "num_tokens": 110074418.0, "step": 2883 }, { "epoch": 0.3668744434550312, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.642586708068848, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8589657545089722, "num_tokens": 110109255.0, "step": 2884 }, { "epoch": 0.3670016537336217, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.482063293457031, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8600091934204102, "num_tokens": 110150180.0, "step": 2885 }, { "epoch": 0.3671288640122122, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.436972618103027, "learning_rate": 1e-06, "loss": 0.5063, "mean_token_accuracy": 0.8377361297607422, "num_tokens": 110193653.0, "step": 2886 }, { "epoch": 0.3672560742908027, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.489669799804688, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8611743450164795, "num_tokens": 110231020.0, "step": 2887 }, { "epoch": 0.36738328456939323, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.456035614013672, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.869577169418335, "num_tokens": 110269397.0, "step": 2888 }, { "epoch": 0.3675104948479837, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.44754409790039, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8687028884887695, "num_tokens": 110311188.0, "step": 2889 }, { "epoch": 0.36763770512657423, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.46635913848877, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.858688473701477, "num_tokens": 110343801.0, "step": 2890 }, { "epoch": 0.36776491540516476, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.448344230651855, "learning_rate": 1e-06, "loss": 0.5027, "mean_token_accuracy": 0.8383461236953735, "num_tokens": 110385830.0, "step": 2891 }, { "epoch": 0.36789212568375523, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.434968948364258, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8536242842674255, "num_tokens": 110421494.0, "step": 2892 }, { "epoch": 0.36801933596234576, "ewc_loss": 0.01416015625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.439972877502441, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8532047271728516, "num_tokens": 110463132.0, "step": 2893 }, { "epoch": 0.3681465462409363, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.436271667480469, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8569296598434448, "num_tokens": 110498330.0, "step": 2894 }, { "epoch": 0.36827375651952676, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.41059398651123, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8624173402786255, "num_tokens": 110534438.0, "step": 2895 }, { "epoch": 0.3684009667981173, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.428047180175781, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.849079430103302, "num_tokens": 110572059.0, "step": 2896 }, { "epoch": 0.3685281770767078, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.412652969360352, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8562934994697571, "num_tokens": 110611012.0, "step": 2897 }, { "epoch": 0.3686553873552983, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.449711799621582, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8618727326393127, "num_tokens": 110647429.0, "step": 2898 }, { "epoch": 0.3687825976338888, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.457412719726562, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8489356637001038, "num_tokens": 110685052.0, "step": 2899 }, { "epoch": 0.36890980791247935, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.463058471679688, "learning_rate": 1e-06, "loss": 0.507, "mean_token_accuracy": 0.833570122718811, "num_tokens": 110725222.0, "step": 2900 }, { "epoch": 0.3690370181910698, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.496085166931152, "learning_rate": 1e-06, "loss": 0.4945, "mean_token_accuracy": 0.8435014486312866, "num_tokens": 110764314.0, "step": 2901 }, { "epoch": 0.36916422846966035, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.463964462280273, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8654612302780151, "num_tokens": 110803585.0, "step": 2902 }, { "epoch": 0.3692914387482509, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.443526268005371, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8489443063735962, "num_tokens": 110846532.0, "step": 2903 }, { "epoch": 0.36941864902684135, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.472355842590332, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8510383367538452, "num_tokens": 110889807.0, "step": 2904 }, { "epoch": 0.3695458593054319, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.467856407165527, "learning_rate": 1e-06, "loss": 0.5105, "mean_token_accuracy": 0.8347917795181274, "num_tokens": 110930567.0, "step": 2905 }, { "epoch": 0.3696730695840224, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.434776306152344, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.860824704170227, "num_tokens": 110968725.0, "step": 2906 }, { "epoch": 0.3698002798626129, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.463512420654297, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.878926694393158, "num_tokens": 111002724.0, "step": 2907 }, { "epoch": 0.3699274901412034, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.49994945526123, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8535454273223877, "num_tokens": 111036668.0, "step": 2908 }, { "epoch": 0.37005470041979394, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.45709228515625, "learning_rate": 1e-06, "loss": 0.4835, "mean_token_accuracy": 0.8452188968658447, "num_tokens": 111077712.0, "step": 2909 }, { "epoch": 0.3701819106983844, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.435927391052246, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.8444979190826416, "num_tokens": 111118856.0, "step": 2910 }, { "epoch": 0.37030912097697494, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.503939628601074, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8450672626495361, "num_tokens": 111156712.0, "step": 2911 }, { "epoch": 0.37043633125556547, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.44375228881836, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8553528785705566, "num_tokens": 111195542.0, "step": 2912 }, { "epoch": 0.37056354153415594, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.488203048706055, "learning_rate": 1e-06, "loss": 0.5248, "mean_token_accuracy": 0.8359802961349487, "num_tokens": 111237827.0, "step": 2913 }, { "epoch": 0.37069075181274647, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.42170238494873, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8619237542152405, "num_tokens": 111278802.0, "step": 2914 }, { "epoch": 0.370817962091337, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.492277145385742, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8514492511749268, "num_tokens": 111313795.0, "step": 2915 }, { "epoch": 0.37094517236992747, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.411725044250488, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8604963421821594, "num_tokens": 111353798.0, "step": 2916 }, { "epoch": 0.371072382648518, "ewc_loss": 0.01416015625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.464494705200195, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8544970750808716, "num_tokens": 111390613.0, "step": 2917 }, { "epoch": 0.3711995929271085, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.458669662475586, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8547232747077942, "num_tokens": 111432043.0, "step": 2918 }, { "epoch": 0.371326803205699, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.446760177612305, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8606449365615845, "num_tokens": 111476575.0, "step": 2919 }, { "epoch": 0.3714540134842895, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.482128143310547, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.854222297668457, "num_tokens": 111515304.0, "step": 2920 }, { "epoch": 0.37158122376288005, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.540315628051758, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8636279106140137, "num_tokens": 111550246.0, "step": 2921 }, { "epoch": 0.3717084340414705, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.471282005310059, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8549870252609253, "num_tokens": 111592189.0, "step": 2922 }, { "epoch": 0.37183564432006105, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.466421127319336, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8577932715415955, "num_tokens": 111638177.0, "step": 2923 }, { "epoch": 0.3719628545986516, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.413457870483398, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8561120629310608, "num_tokens": 111674365.0, "step": 2924 }, { "epoch": 0.37209006487724205, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.512914657592773, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.848772406578064, "num_tokens": 111714571.0, "step": 2925 }, { "epoch": 0.3722172751558326, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.446090698242188, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8549004197120667, "num_tokens": 111755214.0, "step": 2926 }, { "epoch": 0.3723444854344231, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.537744522094727, "learning_rate": 1e-06, "loss": 0.5357, "mean_token_accuracy": 0.8269884586334229, "num_tokens": 111796516.0, "step": 2927 }, { "epoch": 0.3724716957130136, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.516819953918457, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.8396432995796204, "num_tokens": 111832615.0, "step": 2928 }, { "epoch": 0.3725989059916041, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.45114803314209, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8446066379547119, "num_tokens": 111874682.0, "step": 2929 }, { "epoch": 0.37272611627019464, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.500236511230469, "learning_rate": 1e-06, "loss": 0.502, "mean_token_accuracy": 0.8418365120887756, "num_tokens": 111916369.0, "step": 2930 }, { "epoch": 0.37285332654878517, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.536417007446289, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8462622761726379, "num_tokens": 111950671.0, "step": 2931 }, { "epoch": 0.37298053682737564, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.51496696472168, "learning_rate": 1e-06, "loss": 0.5022, "mean_token_accuracy": 0.8360280990600586, "num_tokens": 111988846.0, "step": 2932 }, { "epoch": 0.37310774710596617, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.499191284179688, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.863497257232666, "num_tokens": 112022853.0, "step": 2933 }, { "epoch": 0.3732349573845567, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.490355491638184, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8497251272201538, "num_tokens": 112065631.0, "step": 2934 }, { "epoch": 0.37336216766314717, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.455836296081543, "learning_rate": 1e-06, "loss": 0.4889, "mean_token_accuracy": 0.8428834080696106, "num_tokens": 112107236.0, "step": 2935 }, { "epoch": 0.3734893779417377, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.459137916564941, "learning_rate": 1e-06, "loss": 0.5168, "mean_token_accuracy": 0.832222580909729, "num_tokens": 112153914.0, "step": 2936 }, { "epoch": 0.3736165882203282, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.527549743652344, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8563162684440613, "num_tokens": 112187596.0, "step": 2937 }, { "epoch": 0.3737437984989187, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.504799842834473, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8513363599777222, "num_tokens": 112222409.0, "step": 2938 }, { "epoch": 0.37387100877750923, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.549602508544922, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8413969278335571, "num_tokens": 112257699.0, "step": 2939 }, { "epoch": 0.37399821905609976, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.4886474609375, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8615973591804504, "num_tokens": 112292431.0, "step": 2940 }, { "epoch": 0.37412542933469023, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.47733211517334, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8547884225845337, "num_tokens": 112331241.0, "step": 2941 }, { "epoch": 0.37425263961328076, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.522791862487793, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8498438596725464, "num_tokens": 112368486.0, "step": 2942 }, { "epoch": 0.3743798498918713, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.531572341918945, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8450421094894409, "num_tokens": 112409638.0, "step": 2943 }, { "epoch": 0.37450706017046176, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.464570045471191, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8486509323120117, "num_tokens": 112448257.0, "step": 2944 }, { "epoch": 0.3746342704490523, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.384872436523438, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8678933382034302, "num_tokens": 112488739.0, "step": 2945 }, { "epoch": 0.3747614807276428, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.533868789672852, "learning_rate": 1e-06, "loss": 0.4993, "mean_token_accuracy": 0.8384663462638855, "num_tokens": 112528621.0, "step": 2946 }, { "epoch": 0.3748886910062333, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.518904685974121, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8488262891769409, "num_tokens": 112564615.0, "step": 2947 }, { "epoch": 0.3750159012848238, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.527588844299316, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8428170084953308, "num_tokens": 112598190.0, "step": 2948 }, { "epoch": 0.37514311156341434, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.532855987548828, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.852882444858551, "num_tokens": 112632078.0, "step": 2949 }, { "epoch": 0.3752703218420048, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.519532203674316, "learning_rate": 1e-06, "loss": 0.5264, "mean_token_accuracy": 0.8310805559158325, "num_tokens": 112667848.0, "step": 2950 }, { "epoch": 0.37539753212059535, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.529953956604004, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.851056694984436, "num_tokens": 112707037.0, "step": 2951 }, { "epoch": 0.3755247423991859, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.490889549255371, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8625809550285339, "num_tokens": 112751921.0, "step": 2952 }, { "epoch": 0.37565195267777635, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.50304126739502, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8583104610443115, "num_tokens": 112793061.0, "step": 2953 }, { "epoch": 0.3757791629563669, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.52888011932373, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8636727333068848, "num_tokens": 112833852.0, "step": 2954 }, { "epoch": 0.3759063732349574, "ewc_loss": 0.014404296875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.507593154907227, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8494595289230347, "num_tokens": 112866718.0, "step": 2955 }, { "epoch": 0.3760335835135479, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.505352020263672, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8550366163253784, "num_tokens": 112909420.0, "step": 2956 }, { "epoch": 0.3761607937921384, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.533823013305664, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8564192056655884, "num_tokens": 112944570.0, "step": 2957 }, { "epoch": 0.37628800407072893, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.468737602233887, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8657079339027405, "num_tokens": 112984107.0, "step": 2958 }, { "epoch": 0.3764152143493194, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.45662784576416, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8766660690307617, "num_tokens": 113023066.0, "step": 2959 }, { "epoch": 0.37654242462790993, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.499741554260254, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8511350154876709, "num_tokens": 113060093.0, "step": 2960 }, { "epoch": 0.37666963490650046, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.52641773223877, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8452508449554443, "num_tokens": 113095940.0, "step": 2961 }, { "epoch": 0.37679684518509093, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.565396308898926, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8639416694641113, "num_tokens": 113130699.0, "step": 2962 }, { "epoch": 0.37692405546368146, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.49889850616455, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8534528017044067, "num_tokens": 113169145.0, "step": 2963 }, { "epoch": 0.377051265742272, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.578717231750488, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8684821128845215, "num_tokens": 113201987.0, "step": 2964 }, { "epoch": 0.37717847602086246, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.574918746948242, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8595782518386841, "num_tokens": 113239588.0, "step": 2965 }, { "epoch": 0.377305686299453, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.497276306152344, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8551310300827026, "num_tokens": 113277075.0, "step": 2966 }, { "epoch": 0.3774328965780435, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.51909351348877, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8806477189064026, "num_tokens": 113313708.0, "step": 2967 }, { "epoch": 0.377560106856634, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.526961326599121, "learning_rate": 1e-06, "loss": 0.509, "mean_token_accuracy": 0.835858941078186, "num_tokens": 113344027.0, "step": 2968 }, { "epoch": 0.3776873171352245, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.52498722076416, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8575993776321411, "num_tokens": 113388444.0, "step": 2969 }, { "epoch": 0.37781452741381505, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.477437973022461, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8494003415107727, "num_tokens": 113428660.0, "step": 2970 }, { "epoch": 0.3779417376924055, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.550877571105957, "learning_rate": 1e-06, "loss": 0.5188, "mean_token_accuracy": 0.8327527046203613, "num_tokens": 113469160.0, "step": 2971 }, { "epoch": 0.37806894797099605, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.480295181274414, "learning_rate": 1e-06, "loss": 0.5205, "mean_token_accuracy": 0.8299618363380432, "num_tokens": 113504291.0, "step": 2972 }, { "epoch": 0.3781961582495866, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.548587799072266, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8559634685516357, "num_tokens": 113548085.0, "step": 2973 }, { "epoch": 0.37832336852817705, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.525264739990234, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8586527705192566, "num_tokens": 113581664.0, "step": 2974 }, { "epoch": 0.3784505788067676, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.48934268951416, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8670420050621033, "num_tokens": 113618552.0, "step": 2975 }, { "epoch": 0.3785777890853581, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.591131210327148, "learning_rate": 1e-06, "loss": 0.5265, "mean_token_accuracy": 0.834196150302887, "num_tokens": 113653406.0, "step": 2976 }, { "epoch": 0.3787049993639486, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.542522430419922, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8627142906188965, "num_tokens": 113693803.0, "step": 2977 }, { "epoch": 0.3788322096425391, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.494014739990234, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8685873746871948, "num_tokens": 113734637.0, "step": 2978 }, { "epoch": 0.37895941992112964, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.479968070983887, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8465539216995239, "num_tokens": 113776579.0, "step": 2979 }, { "epoch": 0.37908663019972016, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.498786926269531, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8433433771133423, "num_tokens": 113818857.0, "step": 2980 }, { "epoch": 0.37921384047831064, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.558228492736816, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8453830480575562, "num_tokens": 113853248.0, "step": 2981 }, { "epoch": 0.37934105075690117, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.498200416564941, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8595197200775146, "num_tokens": 113894206.0, "step": 2982 }, { "epoch": 0.3794682610354917, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.5195951461792, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8614435195922852, "num_tokens": 113931309.0, "step": 2983 }, { "epoch": 0.37959547131408217, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.490829467773438, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.842907190322876, "num_tokens": 113976559.0, "step": 2984 }, { "epoch": 0.3797226815926727, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.585009574890137, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8524799346923828, "num_tokens": 114010814.0, "step": 2985 }, { "epoch": 0.3798498918712632, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.518640518188477, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8721826672554016, "num_tokens": 114049389.0, "step": 2986 }, { "epoch": 0.3799771021498537, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.553187370300293, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8713510036468506, "num_tokens": 114081463.0, "step": 2987 }, { "epoch": 0.3801043124284442, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.538570404052734, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8701400756835938, "num_tokens": 114117175.0, "step": 2988 }, { "epoch": 0.38023152270703475, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.566125869750977, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8431047797203064, "num_tokens": 114153764.0, "step": 2989 }, { "epoch": 0.3803587329856252, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.540872573852539, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8534051179885864, "num_tokens": 114191487.0, "step": 2990 }, { "epoch": 0.38048594326421575, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.632863998413086, "learning_rate": 1e-06, "loss": 0.5229, "mean_token_accuracy": 0.8289734125137329, "num_tokens": 114233880.0, "step": 2991 }, { "epoch": 0.3806131535428063, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.520221710205078, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8522272109985352, "num_tokens": 114269238.0, "step": 2992 }, { "epoch": 0.38074036382139675, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.489557266235352, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8657770156860352, "num_tokens": 114307645.0, "step": 2993 }, { "epoch": 0.3808675740999873, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.55632209777832, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8579509258270264, "num_tokens": 114341800.0, "step": 2994 }, { "epoch": 0.3809947843785778, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.464800834655762, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8629377484321594, "num_tokens": 114378363.0, "step": 2995 }, { "epoch": 0.3811219946571683, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.573321342468262, "learning_rate": 1e-06, "loss": 0.4891, "mean_token_accuracy": 0.8421010971069336, "num_tokens": 114414489.0, "step": 2996 }, { "epoch": 0.3812492049357588, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.528497695922852, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8484472632408142, "num_tokens": 114449197.0, "step": 2997 }, { "epoch": 0.38137641521434934, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.518144607543945, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8517995476722717, "num_tokens": 114478654.0, "step": 2998 }, { "epoch": 0.3815036254929398, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.503409385681152, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8628390431404114, "num_tokens": 114521807.0, "step": 2999 }, { "epoch": 0.38163083577153034, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.519874572753906, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8671985864639282, "num_tokens": 114558297.0, "step": 3000 }, { "epoch": 0.38175804605012087, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.496201515197754, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8602795004844666, "num_tokens": 114599630.0, "step": 3001 }, { "epoch": 0.38188525632871134, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.554620742797852, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8514953851699829, "num_tokens": 114638697.0, "step": 3002 }, { "epoch": 0.38201246660730187, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.52906608581543, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8553193211555481, "num_tokens": 114676106.0, "step": 3003 }, { "epoch": 0.3821396768858924, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.534096717834473, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.851816713809967, "num_tokens": 114712026.0, "step": 3004 }, { "epoch": 0.38226688716448287, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.56046199798584, "learning_rate": 1e-06, "loss": 0.5165, "mean_token_accuracy": 0.8323365449905396, "num_tokens": 114752348.0, "step": 3005 }, { "epoch": 0.3823940974430734, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.53884220123291, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8583061695098877, "num_tokens": 114789922.0, "step": 3006 }, { "epoch": 0.38252130772166393, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.485014915466309, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.848250150680542, "num_tokens": 114838834.0, "step": 3007 }, { "epoch": 0.3826485180002544, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.553088188171387, "learning_rate": 1e-06, "loss": 0.5062, "mean_token_accuracy": 0.8335721492767334, "num_tokens": 114873320.0, "step": 3008 }, { "epoch": 0.38277572827884493, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.52825927734375, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8658579587936401, "num_tokens": 114908293.0, "step": 3009 }, { "epoch": 0.38290293855743546, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.442851066589355, "learning_rate": 1e-06, "loss": 0.4815, "mean_token_accuracy": 0.844502329826355, "num_tokens": 114950083.0, "step": 3010 }, { "epoch": 0.38303014883602593, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.61906909942627, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8585333824157715, "num_tokens": 114982353.0, "step": 3011 }, { "epoch": 0.38315735911461646, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.554240226745605, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8593591451644897, "num_tokens": 115017668.0, "step": 3012 }, { "epoch": 0.383284569393207, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.486116409301758, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8652712106704712, "num_tokens": 115055266.0, "step": 3013 }, { "epoch": 0.38341177967179746, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.588082313537598, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8492909073829651, "num_tokens": 115089728.0, "step": 3014 }, { "epoch": 0.383538989950388, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.48598575592041, "learning_rate": 1e-06, "loss": 0.4925, "mean_token_accuracy": 0.8378614187240601, "num_tokens": 115135192.0, "step": 3015 }, { "epoch": 0.3836662002289785, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.604640007019043, "learning_rate": 1e-06, "loss": 0.5092, "mean_token_accuracy": 0.8366105556488037, "num_tokens": 115170350.0, "step": 3016 }, { "epoch": 0.383793410507569, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.504756927490234, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8601135015487671, "num_tokens": 115202724.0, "step": 3017 }, { "epoch": 0.3839206207861595, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.514318466186523, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8537439107894897, "num_tokens": 115239274.0, "step": 3018 }, { "epoch": 0.38404783106475004, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.559643745422363, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8618203401565552, "num_tokens": 115278330.0, "step": 3019 }, { "epoch": 0.3841750413433405, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.531280517578125, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8482877016067505, "num_tokens": 115321258.0, "step": 3020 }, { "epoch": 0.38430225162193105, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.529151916503906, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8553866147994995, "num_tokens": 115362609.0, "step": 3021 }, { "epoch": 0.3844294619005216, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.572295188903809, "learning_rate": 1e-06, "loss": 0.5425, "mean_token_accuracy": 0.8275717496871948, "num_tokens": 115405814.0, "step": 3022 }, { "epoch": 0.38455667217911205, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.563088417053223, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8474640846252441, "num_tokens": 115440856.0, "step": 3023 }, { "epoch": 0.3846838824577026, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.479976654052734, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8600924015045166, "num_tokens": 115482582.0, "step": 3024 }, { "epoch": 0.3848110927362931, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.578353881835938, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8447131514549255, "num_tokens": 115527268.0, "step": 3025 }, { "epoch": 0.3849383030148836, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.55699348449707, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8497287034988403, "num_tokens": 115563757.0, "step": 3026 }, { "epoch": 0.3850655132934741, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.491462707519531, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8582733869552612, "num_tokens": 115599629.0, "step": 3027 }, { "epoch": 0.38519272357206463, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.596308708190918, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8582226037979126, "num_tokens": 115634261.0, "step": 3028 }, { "epoch": 0.3853199338506551, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.529281616210938, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8664474487304688, "num_tokens": 115671150.0, "step": 3029 }, { "epoch": 0.38544714412924563, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.50417709350586, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8671647310256958, "num_tokens": 115714328.0, "step": 3030 }, { "epoch": 0.38557435440783616, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.55381965637207, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8690251111984253, "num_tokens": 115747118.0, "step": 3031 }, { "epoch": 0.3857015646864267, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.511677742004395, "learning_rate": 1e-06, "loss": 0.5153, "mean_token_accuracy": 0.834307849407196, "num_tokens": 115790309.0, "step": 3032 }, { "epoch": 0.38582877496501716, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.520048141479492, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8663231730461121, "num_tokens": 115825630.0, "step": 3033 }, { "epoch": 0.3859559852436077, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.573041915893555, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8618701100349426, "num_tokens": 115858848.0, "step": 3034 }, { "epoch": 0.3860831955221982, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.537348747253418, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8636260628700256, "num_tokens": 115900024.0, "step": 3035 }, { "epoch": 0.3862104058007887, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.528432846069336, "learning_rate": 1e-06, "loss": 0.528, "mean_token_accuracy": 0.8296822309494019, "num_tokens": 115938899.0, "step": 3036 }, { "epoch": 0.3863376160793792, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.52338981628418, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8621598482131958, "num_tokens": 115972199.0, "step": 3037 }, { "epoch": 0.38646482635796975, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.540738105773926, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8642504215240479, "num_tokens": 116011506.0, "step": 3038 }, { "epoch": 0.3865920366365602, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.598130226135254, "learning_rate": 1e-06, "loss": 0.5272, "mean_token_accuracy": 0.8316367268562317, "num_tokens": 116051428.0, "step": 3039 }, { "epoch": 0.38671924691515075, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.518317222595215, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8632745742797852, "num_tokens": 116090307.0, "step": 3040 }, { "epoch": 0.3868464571937413, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.57125473022461, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8447264432907104, "num_tokens": 116123381.0, "step": 3041 }, { "epoch": 0.38697366747233175, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.541413307189941, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8573347330093384, "num_tokens": 116157755.0, "step": 3042 }, { "epoch": 0.3871008777509223, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.520673751831055, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8547817468643188, "num_tokens": 116197371.0, "step": 3043 }, { "epoch": 0.3872280880295128, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.549240112304688, "learning_rate": 1e-06, "loss": 0.5037, "mean_token_accuracy": 0.8395575284957886, "num_tokens": 116240544.0, "step": 3044 }, { "epoch": 0.3873552983081033, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.581066131591797, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8515892624855042, "num_tokens": 116272079.0, "step": 3045 }, { "epoch": 0.3874825085866938, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.509058952331543, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8510093688964844, "num_tokens": 116309641.0, "step": 3046 }, { "epoch": 0.38760971886528434, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.52766227722168, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8488377332687378, "num_tokens": 116344470.0, "step": 3047 }, { "epoch": 0.3877369291438748, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.566045761108398, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8584887981414795, "num_tokens": 116380177.0, "step": 3048 }, { "epoch": 0.38786413942246534, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.574067115783691, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8648848533630371, "num_tokens": 116416482.0, "step": 3049 }, { "epoch": 0.38799134970105587, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.541072845458984, "learning_rate": 1e-06, "loss": 0.4971, "mean_token_accuracy": 0.8412352800369263, "num_tokens": 116456103.0, "step": 3050 }, { "epoch": 0.38811855997964634, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.550670623779297, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8599569797515869, "num_tokens": 116496810.0, "step": 3051 }, { "epoch": 0.38824577025823687, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.572295188903809, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8511848449707031, "num_tokens": 116532999.0, "step": 3052 }, { "epoch": 0.3883729805368274, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.590241432189941, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8623272180557251, "num_tokens": 116565700.0, "step": 3053 }, { "epoch": 0.38850019081541787, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.548171043395996, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8586001396179199, "num_tokens": 116602798.0, "step": 3054 }, { "epoch": 0.3886274010940084, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.595054626464844, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.864727258682251, "num_tokens": 116638743.0, "step": 3055 }, { "epoch": 0.3887546113725989, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.56679916381836, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8671749234199524, "num_tokens": 116672290.0, "step": 3056 }, { "epoch": 0.3888818216511894, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.540335655212402, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8558850884437561, "num_tokens": 116712593.0, "step": 3057 }, { "epoch": 0.3890090319297799, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.583006858825684, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8527294397354126, "num_tokens": 116748663.0, "step": 3058 }, { "epoch": 0.38913624220837045, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.51901912689209, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8510684370994568, "num_tokens": 116796018.0, "step": 3059 }, { "epoch": 0.3892634524869609, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.588630676269531, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8522601127624512, "num_tokens": 116837287.0, "step": 3060 }, { "epoch": 0.38939066276555145, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.598657608032227, "learning_rate": 1e-06, "loss": 0.5199, "mean_token_accuracy": 0.8387026786804199, "num_tokens": 116872163.0, "step": 3061 }, { "epoch": 0.389517873044142, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.527524948120117, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8626576662063599, "num_tokens": 116913445.0, "step": 3062 }, { "epoch": 0.38964508332273246, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.499459266662598, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.8450442552566528, "num_tokens": 116954469.0, "step": 3063 }, { "epoch": 0.389772293601323, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.56246280670166, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8540024161338806, "num_tokens": 116994339.0, "step": 3064 }, { "epoch": 0.3898995038799135, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.568769454956055, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8664125800132751, "num_tokens": 117030867.0, "step": 3065 }, { "epoch": 0.390026714158504, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.59163761138916, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8630638122558594, "num_tokens": 117068309.0, "step": 3066 }, { "epoch": 0.3901539244370945, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.533713340759277, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.850294828414917, "num_tokens": 117105278.0, "step": 3067 }, { "epoch": 0.39028113471568504, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.609496116638184, "learning_rate": 1e-06, "loss": 0.489, "mean_token_accuracy": 0.84633469581604, "num_tokens": 117143096.0, "step": 3068 }, { "epoch": 0.3904083449942755, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.555017471313477, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8545582890510559, "num_tokens": 117179638.0, "step": 3069 }, { "epoch": 0.39053555527286604, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.618301391601562, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8667604923248291, "num_tokens": 117211560.0, "step": 3070 }, { "epoch": 0.39066276555145657, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.6162109375, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.846077024936676, "num_tokens": 117250622.0, "step": 3071 }, { "epoch": 0.39078997583004704, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.503680229187012, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8743811845779419, "num_tokens": 117287600.0, "step": 3072 }, { "epoch": 0.39091718610863757, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.580244064331055, "learning_rate": 1e-06, "loss": 0.4984, "mean_token_accuracy": 0.839790940284729, "num_tokens": 117328188.0, "step": 3073 }, { "epoch": 0.3910443963872281, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.556304931640625, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8633922338485718, "num_tokens": 117367282.0, "step": 3074 }, { "epoch": 0.39117160666581857, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.558357238769531, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8666144013404846, "num_tokens": 117403376.0, "step": 3075 }, { "epoch": 0.3912988169444091, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.588411331176758, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.8458577394485474, "num_tokens": 117438661.0, "step": 3076 }, { "epoch": 0.39142602722299963, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.5979585647583, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8592342734336853, "num_tokens": 117477351.0, "step": 3077 }, { "epoch": 0.3915532375015901, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.543932914733887, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8521358966827393, "num_tokens": 117512773.0, "step": 3078 }, { "epoch": 0.39168044778018063, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.612467765808105, "learning_rate": 1e-06, "loss": 0.4991, "mean_token_accuracy": 0.8401644825935364, "num_tokens": 117555370.0, "step": 3079 }, { "epoch": 0.39180765805877116, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.51669692993164, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8621310591697693, "num_tokens": 117594208.0, "step": 3080 }, { "epoch": 0.3919348683373617, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.551231384277344, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8583136796951294, "num_tokens": 117632715.0, "step": 3081 }, { "epoch": 0.39206207861595216, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.602559089660645, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.85040682554245, "num_tokens": 117669221.0, "step": 3082 }, { "epoch": 0.3921892888945427, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.522850036621094, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8489289283752441, "num_tokens": 117709559.0, "step": 3083 }, { "epoch": 0.3923164991731332, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.60774040222168, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.862027108669281, "num_tokens": 117745240.0, "step": 3084 }, { "epoch": 0.3924437094517237, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.504849433898926, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8566933870315552, "num_tokens": 117787897.0, "step": 3085 }, { "epoch": 0.3925709197303142, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.5881986618042, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8509451150894165, "num_tokens": 117826862.0, "step": 3086 }, { "epoch": 0.39269813000890474, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.582635879516602, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.850223183631897, "num_tokens": 117866932.0, "step": 3087 }, { "epoch": 0.3928253402874952, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.49423885345459, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8556994199752808, "num_tokens": 117905538.0, "step": 3088 }, { "epoch": 0.39295255056608575, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.541814804077148, "learning_rate": 1e-06, "loss": 0.4925, "mean_token_accuracy": 0.8434343934059143, "num_tokens": 117945123.0, "step": 3089 }, { "epoch": 0.3930797608446763, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.569165229797363, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8574681878089905, "num_tokens": 117988743.0, "step": 3090 }, { "epoch": 0.39320697112326675, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.520392417907715, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8690158724784851, "num_tokens": 118026078.0, "step": 3091 }, { "epoch": 0.3933341814018573, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.552603721618652, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8668678998947144, "num_tokens": 118060105.0, "step": 3092 }, { "epoch": 0.3934613916804478, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.570934295654297, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8820040225982666, "num_tokens": 118089677.0, "step": 3093 }, { "epoch": 0.3935886019590383, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.518159866333008, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8677233457565308, "num_tokens": 118128449.0, "step": 3094 }, { "epoch": 0.3937158122376288, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.611023902893066, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8636474609375, "num_tokens": 118168411.0, "step": 3095 }, { "epoch": 0.39384302251621933, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.538047790527344, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8659734725952148, "num_tokens": 118206097.0, "step": 3096 }, { "epoch": 0.3939702327948098, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.607667922973633, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.868169903755188, "num_tokens": 118244429.0, "step": 3097 }, { "epoch": 0.39409744307340033, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.55172061920166, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8612450957298279, "num_tokens": 118279125.0, "step": 3098 }, { "epoch": 0.39422465335199086, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.617783546447754, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8529130220413208, "num_tokens": 118313156.0, "step": 3099 }, { "epoch": 0.39435186363058133, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.563847541809082, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.8417431712150574, "num_tokens": 118349868.0, "step": 3100 }, { "epoch": 0.39447907390917186, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.573185920715332, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8461897969245911, "num_tokens": 118394998.0, "step": 3101 }, { "epoch": 0.3946062841877624, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.575424194335938, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8673015236854553, "num_tokens": 118436002.0, "step": 3102 }, { "epoch": 0.39473349446635286, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.55482292175293, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8603689670562744, "num_tokens": 118475094.0, "step": 3103 }, { "epoch": 0.3948607047449434, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.554391860961914, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8529757857322693, "num_tokens": 118514164.0, "step": 3104 }, { "epoch": 0.3949879150235339, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.604981422424316, "learning_rate": 1e-06, "loss": 0.4728, "mean_token_accuracy": 0.8507766723632812, "num_tokens": 118554271.0, "step": 3105 }, { "epoch": 0.3951151253021244, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.557812690734863, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8559116721153259, "num_tokens": 118597123.0, "step": 3106 }, { "epoch": 0.3952423355807149, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.56437873840332, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8495916128158569, "num_tokens": 118636683.0, "step": 3107 }, { "epoch": 0.39536954585930545, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.614974021911621, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8585353493690491, "num_tokens": 118665770.0, "step": 3108 }, { "epoch": 0.3954967561378959, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.621901512145996, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8666204214096069, "num_tokens": 118701652.0, "step": 3109 }, { "epoch": 0.39562396641648645, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.558475494384766, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8617088794708252, "num_tokens": 118741756.0, "step": 3110 }, { "epoch": 0.395751176695077, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.559124946594238, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8567425012588501, "num_tokens": 118784320.0, "step": 3111 }, { "epoch": 0.39587838697366745, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.632740020751953, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8498910665512085, "num_tokens": 118821456.0, "step": 3112 }, { "epoch": 0.396005597252258, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.61440658569336, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8534805178642273, "num_tokens": 118861746.0, "step": 3113 }, { "epoch": 0.3961328075308485, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.588876724243164, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8608346581459045, "num_tokens": 118900346.0, "step": 3114 }, { "epoch": 0.396260017809439, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.618093490600586, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8566383719444275, "num_tokens": 118939707.0, "step": 3115 }, { "epoch": 0.3963872280880295, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.627552032470703, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8553558588027954, "num_tokens": 118971154.0, "step": 3116 }, { "epoch": 0.39651443836662004, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.638568878173828, "learning_rate": 1e-06, "loss": 0.5058, "mean_token_accuracy": 0.841032862663269, "num_tokens": 119012344.0, "step": 3117 }, { "epoch": 0.3966416486452105, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.635140419006348, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8548676371574402, "num_tokens": 119047677.0, "step": 3118 }, { "epoch": 0.39676885892380104, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.575006484985352, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8699080944061279, "num_tokens": 119086357.0, "step": 3119 }, { "epoch": 0.39689606920239157, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.642070770263672, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8622429966926575, "num_tokens": 119122854.0, "step": 3120 }, { "epoch": 0.39702327948098204, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.582819938659668, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8625097274780273, "num_tokens": 119155449.0, "step": 3121 }, { "epoch": 0.39715048975957257, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.648273468017578, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8504347205162048, "num_tokens": 119197287.0, "step": 3122 }, { "epoch": 0.3972777000381631, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.55666732788086, "learning_rate": 1e-06, "loss": 0.4823, "mean_token_accuracy": 0.8438368439674377, "num_tokens": 119235749.0, "step": 3123 }, { "epoch": 0.39740491031675357, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.671589851379395, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8518610596656799, "num_tokens": 119267282.0, "step": 3124 }, { "epoch": 0.3975321205953441, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.629955291748047, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8473132848739624, "num_tokens": 119304387.0, "step": 3125 }, { "epoch": 0.3976593308739346, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.618584632873535, "learning_rate": 1e-06, "loss": 0.4877, "mean_token_accuracy": 0.8458895683288574, "num_tokens": 119341470.0, "step": 3126 }, { "epoch": 0.3977865411525251, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.607961654663086, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8444388508796692, "num_tokens": 119381067.0, "step": 3127 }, { "epoch": 0.3979137514311156, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.671113014221191, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8549940586090088, "num_tokens": 119415680.0, "step": 3128 }, { "epoch": 0.39804096170970615, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.665231704711914, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8460307717323303, "num_tokens": 119454300.0, "step": 3129 }, { "epoch": 0.3981681719882967, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.661121368408203, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8678603172302246, "num_tokens": 119485599.0, "step": 3130 }, { "epoch": 0.39829538226688715, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.62727165222168, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.840628981590271, "num_tokens": 119525820.0, "step": 3131 }, { "epoch": 0.3984225925454777, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.614408493041992, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8610873222351074, "num_tokens": 119568216.0, "step": 3132 }, { "epoch": 0.3985498028240682, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.645588874816895, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8421680331230164, "num_tokens": 119602914.0, "step": 3133 }, { "epoch": 0.3986770131026587, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.63539981842041, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8514386415481567, "num_tokens": 119637024.0, "step": 3134 }, { "epoch": 0.3988042233812492, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.633874893188477, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.8459023833274841, "num_tokens": 119679265.0, "step": 3135 }, { "epoch": 0.39893143365983974, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.620981216430664, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.845894992351532, "num_tokens": 119716254.0, "step": 3136 }, { "epoch": 0.3990586439384302, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.587275505065918, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8474916219711304, "num_tokens": 119761325.0, "step": 3137 }, { "epoch": 0.39918585421702074, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.674426078796387, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8783568143844604, "num_tokens": 119795468.0, "step": 3138 }, { "epoch": 0.39931306449561127, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.587766647338867, "learning_rate": 1e-06, "loss": 0.4925, "mean_token_accuracy": 0.8455219268798828, "num_tokens": 119833804.0, "step": 3139 }, { "epoch": 0.39944027477420174, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.688285827636719, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8507059812545776, "num_tokens": 119870852.0, "step": 3140 }, { "epoch": 0.39956748505279227, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.615856170654297, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8578193783760071, "num_tokens": 119909258.0, "step": 3141 }, { "epoch": 0.3996946953313828, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.62452507019043, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.843017578125, "num_tokens": 119940701.0, "step": 3142 }, { "epoch": 0.39982190560997327, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.630409240722656, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8552963137626648, "num_tokens": 119976780.0, "step": 3143 }, { "epoch": 0.3999491158885638, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.643510818481445, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8466134667396545, "num_tokens": 120017543.0, "step": 3144 }, { "epoch": 0.40007632616715433, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.64118480682373, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8522286415100098, "num_tokens": 120050036.0, "step": 3145 }, { "epoch": 0.4002035364457448, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.624107360839844, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8690111041069031, "num_tokens": 120084582.0, "step": 3146 }, { "epoch": 0.40033074672433533, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.636245727539062, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8509743809700012, "num_tokens": 120120218.0, "step": 3147 }, { "epoch": 0.40045795700292586, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.533921241760254, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.876788854598999, "num_tokens": 120161650.0, "step": 3148 }, { "epoch": 0.40058516728151633, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.629947662353516, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.863868236541748, "num_tokens": 120197409.0, "step": 3149 }, { "epoch": 0.40071237756010686, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.644573211669922, "learning_rate": 1e-06, "loss": 0.4919, "mean_token_accuracy": 0.8424316644668579, "num_tokens": 120240700.0, "step": 3150 }, { "epoch": 0.4008395878386974, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.5910005569458, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8454556465148926, "num_tokens": 120286741.0, "step": 3151 }, { "epoch": 0.40096679811728786, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.665660858154297, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.857311487197876, "num_tokens": 120323236.0, "step": 3152 }, { "epoch": 0.4010940083958784, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.580267906188965, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8546372652053833, "num_tokens": 120362185.0, "step": 3153 }, { "epoch": 0.4012212186744689, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.647544860839844, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8458471298217773, "num_tokens": 120397524.0, "step": 3154 }, { "epoch": 0.4013484289530594, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.604543685913086, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8587954640388489, "num_tokens": 120439723.0, "step": 3155 }, { "epoch": 0.4014756392316499, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.62060260772705, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8581898212432861, "num_tokens": 120483693.0, "step": 3156 }, { "epoch": 0.40160284951024044, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.556325912475586, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8645615577697754, "num_tokens": 120520558.0, "step": 3157 }, { "epoch": 0.4017300597888309, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.600423812866211, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.864402174949646, "num_tokens": 120557901.0, "step": 3158 }, { "epoch": 0.40185727006742145, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.625272750854492, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8608137369155884, "num_tokens": 120593851.0, "step": 3159 }, { "epoch": 0.401984480346012, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.676250457763672, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.847176194190979, "num_tokens": 120624533.0, "step": 3160 }, { "epoch": 0.40211169062460245, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.577433586120605, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8500350713729858, "num_tokens": 120660820.0, "step": 3161 }, { "epoch": 0.402238900903193, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.558282852172852, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8665936589241028, "num_tokens": 120701485.0, "step": 3162 }, { "epoch": 0.4023661111817835, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.610846519470215, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8646236658096313, "num_tokens": 120736677.0, "step": 3163 }, { "epoch": 0.402493321460374, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.676104545593262, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8528653383255005, "num_tokens": 120773933.0, "step": 3164 }, { "epoch": 0.4026205317389645, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.553812026977539, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.860432505607605, "num_tokens": 120804671.0, "step": 3165 }, { "epoch": 0.40274774201755503, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.663799285888672, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8595941066741943, "num_tokens": 120837042.0, "step": 3166 }, { "epoch": 0.4028749522961455, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.641899108886719, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8519388437271118, "num_tokens": 120875127.0, "step": 3167 }, { "epoch": 0.40300216257473603, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.633340835571289, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8536819219589233, "num_tokens": 120915937.0, "step": 3168 }, { "epoch": 0.40312937285332656, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.700499534606934, "learning_rate": 1e-06, "loss": 0.4847, "mean_token_accuracy": 0.8398628830909729, "num_tokens": 120953334.0, "step": 3169 }, { "epoch": 0.40325658313191703, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.567134857177734, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8549378514289856, "num_tokens": 120986674.0, "step": 3170 }, { "epoch": 0.40338379341050756, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.647395133972168, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8616479635238647, "num_tokens": 121026305.0, "step": 3171 }, { "epoch": 0.4035110036890981, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.641439437866211, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8696757555007935, "num_tokens": 121061644.0, "step": 3172 }, { "epoch": 0.40363821396768856, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.5706205368042, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8512234687805176, "num_tokens": 121102860.0, "step": 3173 }, { "epoch": 0.4037654242462791, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.68275260925293, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8691367506980896, "num_tokens": 121136566.0, "step": 3174 }, { "epoch": 0.4038926345248696, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.670426368713379, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8656575083732605, "num_tokens": 121173759.0, "step": 3175 }, { "epoch": 0.4040198448034601, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.587367057800293, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8554957509040833, "num_tokens": 121219881.0, "step": 3176 }, { "epoch": 0.4041470550820506, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.759669303894043, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8540822267532349, "num_tokens": 121258719.0, "step": 3177 }, { "epoch": 0.40427426536064115, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.638907432556152, "learning_rate": 1e-06, "loss": 0.5129, "mean_token_accuracy": 0.8323988318443298, "num_tokens": 121298013.0, "step": 3178 }, { "epoch": 0.4044014756392316, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.630852699279785, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8674204349517822, "num_tokens": 121329615.0, "step": 3179 }, { "epoch": 0.40452868591782215, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.657878875732422, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8651368618011475, "num_tokens": 121365199.0, "step": 3180 }, { "epoch": 0.4046558961964127, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.615942001342773, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.851478099822998, "num_tokens": 121402393.0, "step": 3181 }, { "epoch": 0.4047831064750032, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.67816162109375, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8548527956008911, "num_tokens": 121439678.0, "step": 3182 }, { "epoch": 0.4049103167535937, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.579608917236328, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8498235940933228, "num_tokens": 121480781.0, "step": 3183 }, { "epoch": 0.4050375270321842, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.69050121307373, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8626217842102051, "num_tokens": 121517614.0, "step": 3184 }, { "epoch": 0.40516473731077474, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.66500186920166, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8534167408943176, "num_tokens": 121552225.0, "step": 3185 }, { "epoch": 0.4052919475893652, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.650519371032715, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.850483775138855, "num_tokens": 121587177.0, "step": 3186 }, { "epoch": 0.40541915786795574, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.621960639953613, "learning_rate": 1e-06, "loss": 0.4987, "mean_token_accuracy": 0.8380364775657654, "num_tokens": 121619547.0, "step": 3187 }, { "epoch": 0.40554636814654627, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.691994667053223, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8626163005828857, "num_tokens": 121657396.0, "step": 3188 }, { "epoch": 0.40567357842513674, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.603164672851562, "learning_rate": 1e-06, "loss": 0.4924, "mean_token_accuracy": 0.838947057723999, "num_tokens": 121691867.0, "step": 3189 }, { "epoch": 0.40580078870372727, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.65127944946289, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8493571281433105, "num_tokens": 121723546.0, "step": 3190 }, { "epoch": 0.4059279989823178, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.56993579864502, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8549189567565918, "num_tokens": 121764589.0, "step": 3191 }, { "epoch": 0.40605520926090827, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.636013984680176, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.8444796204566956, "num_tokens": 121803281.0, "step": 3192 }, { "epoch": 0.4061824195394988, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.660868644714355, "learning_rate": 1e-06, "loss": 0.4894, "mean_token_accuracy": 0.8422012329101562, "num_tokens": 121841710.0, "step": 3193 }, { "epoch": 0.4063096298180893, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.561017036437988, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8535207509994507, "num_tokens": 121881979.0, "step": 3194 }, { "epoch": 0.4064368400966798, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.578771591186523, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.860047459602356, "num_tokens": 121921599.0, "step": 3195 }, { "epoch": 0.4065640503752703, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.626300811767578, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8642420172691345, "num_tokens": 121960545.0, "step": 3196 }, { "epoch": 0.40669126065386085, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.565576553344727, "learning_rate": 1e-06, "loss": 0.5071, "mean_token_accuracy": 0.8373415470123291, "num_tokens": 121999259.0, "step": 3197 }, { "epoch": 0.4068184709324513, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.566957473754883, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8686025142669678, "num_tokens": 122039066.0, "step": 3198 }, { "epoch": 0.40694568121104185, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.584332466125488, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8630425930023193, "num_tokens": 122085035.0, "step": 3199 }, { "epoch": 0.4070728914896324, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.556075096130371, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8696252107620239, "num_tokens": 122128007.0, "step": 3200 }, { "epoch": 0.40720010176822286, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.609647750854492, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8527793884277344, "num_tokens": 122169623.0, "step": 3201 }, { "epoch": 0.4073273120468134, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.645515441894531, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8504693508148193, "num_tokens": 122201755.0, "step": 3202 }, { "epoch": 0.4074545223254039, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.606100082397461, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8661126494407654, "num_tokens": 122233680.0, "step": 3203 }, { "epoch": 0.4075817326039944, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.696975708007812, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8524371981620789, "num_tokens": 122265107.0, "step": 3204 }, { "epoch": 0.4077089428825849, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.611995697021484, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8602544069290161, "num_tokens": 122302750.0, "step": 3205 }, { "epoch": 0.40783615316117544, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.620615005493164, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8467458486557007, "num_tokens": 122346820.0, "step": 3206 }, { "epoch": 0.4079633634397659, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.619680404663086, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8604221343994141, "num_tokens": 122383254.0, "step": 3207 }, { "epoch": 0.40809057371835644, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.644865989685059, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8674947023391724, "num_tokens": 122416693.0, "step": 3208 }, { "epoch": 0.40821778399694697, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.669923782348633, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8604738712310791, "num_tokens": 122452403.0, "step": 3209 }, { "epoch": 0.40834499427553744, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.573838233947754, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8451382517814636, "num_tokens": 122489685.0, "step": 3210 }, { "epoch": 0.40847220455412797, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.607165336608887, "learning_rate": 1e-06, "loss": 0.5003, "mean_token_accuracy": 0.8376253843307495, "num_tokens": 122529079.0, "step": 3211 }, { "epoch": 0.4085994148327185, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.62697982788086, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8485680818557739, "num_tokens": 122569933.0, "step": 3212 }, { "epoch": 0.40872662511130897, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.63003921508789, "learning_rate": 1e-06, "loss": 0.5079, "mean_token_accuracy": 0.8362681865692139, "num_tokens": 122607884.0, "step": 3213 }, { "epoch": 0.4088538353898995, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.594189643859863, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8568024635314941, "num_tokens": 122651968.0, "step": 3214 }, { "epoch": 0.40898104566849003, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.615448951721191, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8601011633872986, "num_tokens": 122688680.0, "step": 3215 }, { "epoch": 0.4091082559470805, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.604007720947266, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8616738319396973, "num_tokens": 122728271.0, "step": 3216 }, { "epoch": 0.40923546622567103, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.653870582580566, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8545735478401184, "num_tokens": 122759691.0, "step": 3217 }, { "epoch": 0.40936267650426156, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.588748931884766, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8543784022331238, "num_tokens": 122800185.0, "step": 3218 }, { "epoch": 0.40948988678285203, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.594852447509766, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.865472674369812, "num_tokens": 122839548.0, "step": 3219 }, { "epoch": 0.40961709706144256, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.633149147033691, "learning_rate": 1e-06, "loss": 0.4832, "mean_token_accuracy": 0.8418748378753662, "num_tokens": 122879621.0, "step": 3220 }, { "epoch": 0.4097443073400331, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.580083847045898, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8507876396179199, "num_tokens": 122921495.0, "step": 3221 }, { "epoch": 0.40987151761862356, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.598047256469727, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.862247884273529, "num_tokens": 122957197.0, "step": 3222 }, { "epoch": 0.4099987278972141, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.64621353149414, "learning_rate": 1e-06, "loss": 0.5011, "mean_token_accuracy": 0.837435245513916, "num_tokens": 122999999.0, "step": 3223 }, { "epoch": 0.4101259381758046, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.612127304077148, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8601806163787842, "num_tokens": 123040402.0, "step": 3224 }, { "epoch": 0.4102531484543951, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.65316104888916, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8602088689804077, "num_tokens": 123080086.0, "step": 3225 }, { "epoch": 0.4103803587329856, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.577128410339355, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8572744131088257, "num_tokens": 123120058.0, "step": 3226 }, { "epoch": 0.41050756901157615, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.628323554992676, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8578060865402222, "num_tokens": 123162003.0, "step": 3227 }, { "epoch": 0.4106347792901666, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.636186599731445, "learning_rate": 1e-06, "loss": 0.5409, "mean_token_accuracy": 0.8229879140853882, "num_tokens": 123202249.0, "step": 3228 }, { "epoch": 0.41076198956875715, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.644976615905762, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8494623303413391, "num_tokens": 123237401.0, "step": 3229 }, { "epoch": 0.4108891998473477, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.653132438659668, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8719402551651001, "num_tokens": 123269364.0, "step": 3230 }, { "epoch": 0.4110164101259382, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.679676055908203, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8583136200904846, "num_tokens": 123307743.0, "step": 3231 }, { "epoch": 0.4111436204045287, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.607076644897461, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8603249788284302, "num_tokens": 123340102.0, "step": 3232 }, { "epoch": 0.4112708306831192, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.674962997436523, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8510911464691162, "num_tokens": 123374956.0, "step": 3233 }, { "epoch": 0.41139804096170973, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.601387977600098, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8648123741149902, "num_tokens": 123418070.0, "step": 3234 }, { "epoch": 0.4115252512403002, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.657098770141602, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8535733222961426, "num_tokens": 123453183.0, "step": 3235 }, { "epoch": 0.41165246151889073, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.61733627319336, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8558006286621094, "num_tokens": 123491526.0, "step": 3236 }, { "epoch": 0.41177967179748126, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 10.333917617797852, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8574402332305908, "num_tokens": 123529743.0, "step": 3237 }, { "epoch": 0.41190688207607173, "ewc_loss": 0.015625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.62317943572998, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8580614924430847, "num_tokens": 123573402.0, "step": 3238 }, { "epoch": 0.41203409235466226, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.58295726776123, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8530242443084717, "num_tokens": 123604167.0, "step": 3239 }, { "epoch": 0.4121613026332528, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.67655086517334, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8521546721458435, "num_tokens": 123648104.0, "step": 3240 }, { "epoch": 0.41228851291184326, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.606407165527344, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8457448482513428, "num_tokens": 123689678.0, "step": 3241 }, { "epoch": 0.4124157231904338, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.66771125793457, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.863276481628418, "num_tokens": 123729257.0, "step": 3242 }, { "epoch": 0.4125429334690243, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.637041091918945, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8578915596008301, "num_tokens": 123772251.0, "step": 3243 }, { "epoch": 0.4126701437476148, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.67072582244873, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8687734603881836, "num_tokens": 123810345.0, "step": 3244 }, { "epoch": 0.4127973540262053, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.631080627441406, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8525254726409912, "num_tokens": 123855276.0, "step": 3245 }, { "epoch": 0.41292456430479585, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.617603302001953, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8590142726898193, "num_tokens": 123896613.0, "step": 3246 }, { "epoch": 0.4130517745833863, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.628959655761719, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8395235538482666, "num_tokens": 123940713.0, "step": 3247 }, { "epoch": 0.41317898486197685, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.627813339233398, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8452908992767334, "num_tokens": 123981485.0, "step": 3248 }, { "epoch": 0.4133061951405674, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.729752540588379, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8561486601829529, "num_tokens": 124018159.0, "step": 3249 }, { "epoch": 0.41343340541915785, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.688455581665039, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8661049604415894, "num_tokens": 124049911.0, "step": 3250 }, { "epoch": 0.4135606156977484, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.718099594116211, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8767953515052795, "num_tokens": 124084260.0, "step": 3251 }, { "epoch": 0.4136878259763389, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.656811714172363, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8590659499168396, "num_tokens": 124122741.0, "step": 3252 }, { "epoch": 0.4138150362549294, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.591028213500977, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8530983924865723, "num_tokens": 124162793.0, "step": 3253 }, { "epoch": 0.4139422465335199, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.671185493469238, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8492888808250427, "num_tokens": 124206889.0, "step": 3254 }, { "epoch": 0.41406945681211044, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.672908782958984, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.867536723613739, "num_tokens": 124244551.0, "step": 3255 }, { "epoch": 0.4141966670907009, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.6863374710083, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8587077856063843, "num_tokens": 124290485.0, "step": 3256 }, { "epoch": 0.41432387736929144, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.678730010986328, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.862453818321228, "num_tokens": 124323763.0, "step": 3257 }, { "epoch": 0.41445108764788197, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.680246353149414, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.856271505355835, "num_tokens": 124363318.0, "step": 3258 }, { "epoch": 0.41457829792647244, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.688782691955566, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8788092136383057, "num_tokens": 124400845.0, "step": 3259 }, { "epoch": 0.41470550820506297, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.616289138793945, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8614524602890015, "num_tokens": 124442546.0, "step": 3260 }, { "epoch": 0.4148327184836535, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.677189826965332, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8509701490402222, "num_tokens": 124477737.0, "step": 3261 }, { "epoch": 0.41495992876224397, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.670587539672852, "learning_rate": 1e-06, "loss": 0.4861, "mean_token_accuracy": 0.8435523509979248, "num_tokens": 124518081.0, "step": 3262 }, { "epoch": 0.4150871390408345, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.677484512329102, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8452099561691284, "num_tokens": 124556361.0, "step": 3263 }, { "epoch": 0.415214349319425, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.710376739501953, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8548545241355896, "num_tokens": 124588490.0, "step": 3264 }, { "epoch": 0.4153415595980155, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.677277565002441, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8546491265296936, "num_tokens": 124624656.0, "step": 3265 }, { "epoch": 0.415468769876606, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.705134391784668, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8426987528800964, "num_tokens": 124660681.0, "step": 3266 }, { "epoch": 0.41559598015519655, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.677190780639648, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.856804370880127, "num_tokens": 124703448.0, "step": 3267 }, { "epoch": 0.415723190433787, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.62840747833252, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8533883690834045, "num_tokens": 124742078.0, "step": 3268 }, { "epoch": 0.41585040071237755, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.74881362915039, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8429887294769287, "num_tokens": 124779894.0, "step": 3269 }, { "epoch": 0.4159776109909681, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.658153533935547, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8505812287330627, "num_tokens": 124817918.0, "step": 3270 }, { "epoch": 0.41610482126955856, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.73442268371582, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8716917634010315, "num_tokens": 124854839.0, "step": 3271 }, { "epoch": 0.4162320315481491, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.708044052124023, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.8401534557342529, "num_tokens": 124886779.0, "step": 3272 }, { "epoch": 0.4163592418267396, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.654150009155273, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8638817667961121, "num_tokens": 124927919.0, "step": 3273 }, { "epoch": 0.4164864521053301, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.708587646484375, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8532072305679321, "num_tokens": 124966475.0, "step": 3274 }, { "epoch": 0.4166136623839206, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.693422317504883, "learning_rate": 1e-06, "loss": 0.5273, "mean_token_accuracy": 0.8315075635910034, "num_tokens": 125004110.0, "step": 3275 }, { "epoch": 0.41674087266251114, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.662601470947266, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8698062896728516, "num_tokens": 125043164.0, "step": 3276 }, { "epoch": 0.4168680829411016, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.691152572631836, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8589613437652588, "num_tokens": 125084255.0, "step": 3277 }, { "epoch": 0.41699529321969214, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.6682710647583, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8687256574630737, "num_tokens": 125124971.0, "step": 3278 }, { "epoch": 0.41712250349828267, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.746159553527832, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8518862724304199, "num_tokens": 125161470.0, "step": 3279 }, { "epoch": 0.4172497137768732, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.668256759643555, "learning_rate": 1e-06, "loss": 0.4934, "mean_token_accuracy": 0.8399374485015869, "num_tokens": 125199117.0, "step": 3280 }, { "epoch": 0.41737692405546367, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.620802879333496, "learning_rate": 1e-06, "loss": 0.5063, "mean_token_accuracy": 0.837164044380188, "num_tokens": 125238136.0, "step": 3281 }, { "epoch": 0.4175041343340542, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.727432250976562, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.853638768196106, "num_tokens": 125281041.0, "step": 3282 }, { "epoch": 0.41763134461264473, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.703064918518066, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8585964441299438, "num_tokens": 125314739.0, "step": 3283 }, { "epoch": 0.4177585548912352, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.665091514587402, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8491705060005188, "num_tokens": 125353129.0, "step": 3284 }, { "epoch": 0.41788576516982573, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.715980529785156, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.854056715965271, "num_tokens": 125394092.0, "step": 3285 }, { "epoch": 0.41801297544841626, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.663443565368652, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8492401838302612, "num_tokens": 125438262.0, "step": 3286 }, { "epoch": 0.41814018572700673, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.617341995239258, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8539914488792419, "num_tokens": 125477842.0, "step": 3287 }, { "epoch": 0.41826739600559726, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.667274475097656, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8675581812858582, "num_tokens": 125517236.0, "step": 3288 }, { "epoch": 0.4183946062841878, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.667040824890137, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8483201265335083, "num_tokens": 125561598.0, "step": 3289 }, { "epoch": 0.41852181656277826, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.683229446411133, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8532369136810303, "num_tokens": 125599104.0, "step": 3290 }, { "epoch": 0.4186490268413688, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.6854248046875, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8559471368789673, "num_tokens": 125635286.0, "step": 3291 }, { "epoch": 0.4187762371199593, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.690879821777344, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8598375916481018, "num_tokens": 125669534.0, "step": 3292 }, { "epoch": 0.4189034473985498, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.69975757598877, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8547972440719604, "num_tokens": 125707966.0, "step": 3293 }, { "epoch": 0.4190306576771403, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.675389289855957, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.851122260093689, "num_tokens": 125741448.0, "step": 3294 }, { "epoch": 0.41915786795573085, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.662964820861816, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8484188914299011, "num_tokens": 125776887.0, "step": 3295 }, { "epoch": 0.4192850782343213, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.732671737670898, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8601183295249939, "num_tokens": 125807899.0, "step": 3296 }, { "epoch": 0.41941228851291185, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.643619537353516, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8633149862289429, "num_tokens": 125847711.0, "step": 3297 }, { "epoch": 0.4195394987915024, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.7135009765625, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8685473203659058, "num_tokens": 125883248.0, "step": 3298 }, { "epoch": 0.41966670907009285, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.647233963012695, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.8437535762786865, "num_tokens": 125925771.0, "step": 3299 }, { "epoch": 0.4197939193486834, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.731772422790527, "learning_rate": 1e-06, "loss": 0.482, "mean_token_accuracy": 0.8432725667953491, "num_tokens": 125968815.0, "step": 3300 }, { "epoch": 0.4199211296272739, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.653366088867188, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8505754470825195, "num_tokens": 126007761.0, "step": 3301 }, { "epoch": 0.4200483399058644, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.686558723449707, "learning_rate": 1e-06, "loss": 0.5021, "mean_token_accuracy": 0.8366756439208984, "num_tokens": 126047872.0, "step": 3302 }, { "epoch": 0.4201755501844549, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.70136547088623, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.861701488494873, "num_tokens": 126081273.0, "step": 3303 }, { "epoch": 0.42030276046304543, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.704365730285645, "learning_rate": 1e-06, "loss": 0.4971, "mean_token_accuracy": 0.8403689861297607, "num_tokens": 126121590.0, "step": 3304 }, { "epoch": 0.4204299707416359, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.637084007263184, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.8381237387657166, "num_tokens": 126168396.0, "step": 3305 }, { "epoch": 0.42055718102022643, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.800759315490723, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8502714037895203, "num_tokens": 126205220.0, "step": 3306 }, { "epoch": 0.42068439129881696, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.689547538757324, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8485205769538879, "num_tokens": 126239131.0, "step": 3307 }, { "epoch": 0.42081160157740743, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.73071575164795, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8566099405288696, "num_tokens": 126271931.0, "step": 3308 }, { "epoch": 0.42093881185599796, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.711732864379883, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8637868762016296, "num_tokens": 126307730.0, "step": 3309 }, { "epoch": 0.4210660221345885, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.661645889282227, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8582781553268433, "num_tokens": 126344528.0, "step": 3310 }, { "epoch": 0.42119323241317896, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.77736759185791, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8638879656791687, "num_tokens": 126377260.0, "step": 3311 }, { "epoch": 0.4213204426917695, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.681134223937988, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.851834237575531, "num_tokens": 126409723.0, "step": 3312 }, { "epoch": 0.42144765297036, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.746971130371094, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8594307899475098, "num_tokens": 126450929.0, "step": 3313 }, { "epoch": 0.4215748632489505, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.677865982055664, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8648871183395386, "num_tokens": 126486878.0, "step": 3314 }, { "epoch": 0.421702073527541, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.75778579711914, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8431887626647949, "num_tokens": 126522884.0, "step": 3315 }, { "epoch": 0.42182928380613155, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.700675964355469, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8567513823509216, "num_tokens": 126560848.0, "step": 3316 }, { "epoch": 0.421956494084722, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.696467399597168, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8574265837669373, "num_tokens": 126592947.0, "step": 3317 }, { "epoch": 0.42208370436331255, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.63277816772461, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8663713932037354, "num_tokens": 126630631.0, "step": 3318 }, { "epoch": 0.4222109146419031, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.772025108337402, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8515336513519287, "num_tokens": 126673395.0, "step": 3319 }, { "epoch": 0.42233812492049355, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.717588424682617, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8718482255935669, "num_tokens": 126711504.0, "step": 3320 }, { "epoch": 0.4224653351990841, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.638945579528809, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8683449029922485, "num_tokens": 126753217.0, "step": 3321 }, { "epoch": 0.4225925454776746, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.652240753173828, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8614040613174438, "num_tokens": 126792982.0, "step": 3322 }, { "epoch": 0.4227197557562651, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.824199676513672, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8498362302780151, "num_tokens": 126823771.0, "step": 3323 }, { "epoch": 0.4228469660348556, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.666399002075195, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8596937656402588, "num_tokens": 126867137.0, "step": 3324 }, { "epoch": 0.42297417631344614, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.716655731201172, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8614342212677002, "num_tokens": 126902680.0, "step": 3325 }, { "epoch": 0.4231013865920366, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.69854736328125, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8673843741416931, "num_tokens": 126941037.0, "step": 3326 }, { "epoch": 0.42322859687062714, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.688274383544922, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8600193858146667, "num_tokens": 126980009.0, "step": 3327 }, { "epoch": 0.42335580714921767, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.697335243225098, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8685643672943115, "num_tokens": 127020022.0, "step": 3328 }, { "epoch": 0.42348301742780814, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.690411567687988, "learning_rate": 1e-06, "loss": 0.5054, "mean_token_accuracy": 0.8370118737220764, "num_tokens": 127060500.0, "step": 3329 }, { "epoch": 0.42361022770639867, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.715498924255371, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8519721031188965, "num_tokens": 127095886.0, "step": 3330 }, { "epoch": 0.4237374379849892, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.690086364746094, "learning_rate": 1e-06, "loss": 0.4801, "mean_token_accuracy": 0.8463771343231201, "num_tokens": 127134334.0, "step": 3331 }, { "epoch": 0.4238646482635797, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.705388069152832, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8486295342445374, "num_tokens": 127174288.0, "step": 3332 }, { "epoch": 0.4239918585421702, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.680950164794922, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8536292314529419, "num_tokens": 127210311.0, "step": 3333 }, { "epoch": 0.4241190688207607, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.653976440429688, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.876799464225769, "num_tokens": 127252813.0, "step": 3334 }, { "epoch": 0.42424627909935125, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.762468338012695, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8438828587532043, "num_tokens": 127285847.0, "step": 3335 }, { "epoch": 0.4243734893779417, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.673479080200195, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8701925873756409, "num_tokens": 127325360.0, "step": 3336 }, { "epoch": 0.42450069965653225, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.68278694152832, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8593595027923584, "num_tokens": 127366446.0, "step": 3337 }, { "epoch": 0.4246279099351228, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.740571022033691, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8658692836761475, "num_tokens": 127402608.0, "step": 3338 }, { "epoch": 0.42475512021371326, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.723092079162598, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.862235426902771, "num_tokens": 127445355.0, "step": 3339 }, { "epoch": 0.4248823304923038, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.701410293579102, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8510259389877319, "num_tokens": 127480523.0, "step": 3340 }, { "epoch": 0.4250095407708943, "ewc_loss": 0.015625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.75083065032959, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8759421110153198, "num_tokens": 127520787.0, "step": 3341 }, { "epoch": 0.4251367510494848, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.707319259643555, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8520387411117554, "num_tokens": 127565576.0, "step": 3342 }, { "epoch": 0.4252639613280753, "ewc_loss": 0.015625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.686234474182129, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.866898775100708, "num_tokens": 127599144.0, "step": 3343 }, { "epoch": 0.42539117160666584, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.706628799438477, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8747872710227966, "num_tokens": 127637361.0, "step": 3344 }, { "epoch": 0.4255183818852563, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.668678283691406, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8463073968887329, "num_tokens": 127675139.0, "step": 3345 }, { "epoch": 0.42564559216384684, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.74193000793457, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8621983528137207, "num_tokens": 127718822.0, "step": 3346 }, { "epoch": 0.42577280244243737, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.756843566894531, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8610788583755493, "num_tokens": 127763332.0, "step": 3347 }, { "epoch": 0.42590001272102784, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.771493911743164, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8679262399673462, "num_tokens": 127800346.0, "step": 3348 }, { "epoch": 0.42602722299961837, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.659333229064941, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8548918962478638, "num_tokens": 127845142.0, "step": 3349 }, { "epoch": 0.4261544332782089, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.757070541381836, "learning_rate": 1e-06, "loss": 0.5074, "mean_token_accuracy": 0.8398372530937195, "num_tokens": 127881131.0, "step": 3350 }, { "epoch": 0.4262816435567994, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.71145248413086, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8705160617828369, "num_tokens": 127917941.0, "step": 3351 }, { "epoch": 0.4264088538353899, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.709521293640137, "learning_rate": 1e-06, "loss": 0.4895, "mean_token_accuracy": 0.8430483341217041, "num_tokens": 127963345.0, "step": 3352 }, { "epoch": 0.42653606411398043, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.83195972442627, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.865334153175354, "num_tokens": 128004839.0, "step": 3353 }, { "epoch": 0.4266632743925709, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.731618881225586, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8494738340377808, "num_tokens": 128042709.0, "step": 3354 }, { "epoch": 0.42679048467116143, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.725005149841309, "learning_rate": 1e-06, "loss": 0.4914, "mean_token_accuracy": 0.8421288728713989, "num_tokens": 128085069.0, "step": 3355 }, { "epoch": 0.42691769494975196, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.754807472229004, "learning_rate": 1e-06, "loss": 0.4735, "mean_token_accuracy": 0.8512287735939026, "num_tokens": 128124300.0, "step": 3356 }, { "epoch": 0.42704490522834243, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.662529945373535, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8541167378425598, "num_tokens": 128164410.0, "step": 3357 }, { "epoch": 0.42717211550693296, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.67726993560791, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8660639524459839, "num_tokens": 128206524.0, "step": 3358 }, { "epoch": 0.4272993257855235, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.708956718444824, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8474671244621277, "num_tokens": 128246387.0, "step": 3359 }, { "epoch": 0.42742653606411396, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.675023078918457, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8667502999305725, "num_tokens": 128287794.0, "step": 3360 }, { "epoch": 0.4275537463427045, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.660430908203125, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8535717725753784, "num_tokens": 128328467.0, "step": 3361 }, { "epoch": 0.427680956621295, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.71172046661377, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8660566806793213, "num_tokens": 128364319.0, "step": 3362 }, { "epoch": 0.4278081668998855, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.683714866638184, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8518388867378235, "num_tokens": 128403892.0, "step": 3363 }, { "epoch": 0.427935377178476, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.740865707397461, "learning_rate": 1e-06, "loss": 0.5072, "mean_token_accuracy": 0.836774468421936, "num_tokens": 128446760.0, "step": 3364 }, { "epoch": 0.42806258745706655, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.75014591217041, "learning_rate": 1e-06, "loss": 0.5188, "mean_token_accuracy": 0.8318930864334106, "num_tokens": 128485498.0, "step": 3365 }, { "epoch": 0.428189797735657, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.67212963104248, "learning_rate": 1e-06, "loss": 0.4941, "mean_token_accuracy": 0.8450999855995178, "num_tokens": 128527871.0, "step": 3366 }, { "epoch": 0.42831700801424755, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.794692993164062, "learning_rate": 1e-06, "loss": 0.4816, "mean_token_accuracy": 0.8431649208068848, "num_tokens": 128567792.0, "step": 3367 }, { "epoch": 0.4284442182928381, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.721365928649902, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8462508916854858, "num_tokens": 128607217.0, "step": 3368 }, { "epoch": 0.42857142857142855, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.72740364074707, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8613368272781372, "num_tokens": 128642787.0, "step": 3369 }, { "epoch": 0.4286986388500191, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.786242485046387, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8667073249816895, "num_tokens": 128680462.0, "step": 3370 }, { "epoch": 0.4288258491286096, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.721073150634766, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8531419038772583, "num_tokens": 128715101.0, "step": 3371 }, { "epoch": 0.4289530594072001, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.669471740722656, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8483574390411377, "num_tokens": 128749601.0, "step": 3372 }, { "epoch": 0.4290802696857906, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.799756050109863, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8518866896629333, "num_tokens": 128788744.0, "step": 3373 }, { "epoch": 0.42920747996438113, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.678338050842285, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8515422344207764, "num_tokens": 128828102.0, "step": 3374 }, { "epoch": 0.4293346902429716, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.685176849365234, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8510022759437561, "num_tokens": 128863610.0, "step": 3375 }, { "epoch": 0.42946190052156213, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.770081520080566, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8695388436317444, "num_tokens": 128898836.0, "step": 3376 }, { "epoch": 0.42958911080015266, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.694685935974121, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8574934005737305, "num_tokens": 128934821.0, "step": 3377 }, { "epoch": 0.42971632107874314, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.752781867980957, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8540947437286377, "num_tokens": 128966914.0, "step": 3378 }, { "epoch": 0.42984353135733366, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.724679946899414, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8620641231536865, "num_tokens": 129008305.0, "step": 3379 }, { "epoch": 0.4299707416359242, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.719242095947266, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8758606314659119, "num_tokens": 129043331.0, "step": 3380 }, { "epoch": 0.4300979519145147, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.759255409240723, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8613159656524658, "num_tokens": 129083513.0, "step": 3381 }, { "epoch": 0.4302251621931052, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.77755069732666, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.855486273765564, "num_tokens": 129113463.0, "step": 3382 }, { "epoch": 0.4303523724716957, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.719346046447754, "learning_rate": 1e-06, "loss": 0.5072, "mean_token_accuracy": 0.8332089185714722, "num_tokens": 129153080.0, "step": 3383 }, { "epoch": 0.43047958275028625, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.761662483215332, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.85225510597229, "num_tokens": 129193338.0, "step": 3384 }, { "epoch": 0.4306067930288767, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.712459564208984, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8459546566009521, "num_tokens": 129234816.0, "step": 3385 }, { "epoch": 0.43073400330746725, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.791874885559082, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.858565628528595, "num_tokens": 129275228.0, "step": 3386 }, { "epoch": 0.4308612135860578, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.729028701782227, "learning_rate": 1e-06, "loss": 0.5018, "mean_token_accuracy": 0.8372577428817749, "num_tokens": 129309747.0, "step": 3387 }, { "epoch": 0.43098842386464825, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.811203956604004, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8589868545532227, "num_tokens": 129342580.0, "step": 3388 }, { "epoch": 0.4311156341432388, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.802009582519531, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.850940465927124, "num_tokens": 129376599.0, "step": 3389 }, { "epoch": 0.4312428444218293, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.703765869140625, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.854428768157959, "num_tokens": 129414783.0, "step": 3390 }, { "epoch": 0.4313700547004198, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.756946563720703, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8551162481307983, "num_tokens": 129450287.0, "step": 3391 }, { "epoch": 0.4314972649790103, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.773008346557617, "learning_rate": 1e-06, "loss": 0.4995, "mean_token_accuracy": 0.8380179405212402, "num_tokens": 129486733.0, "step": 3392 }, { "epoch": 0.43162447525760084, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.765061378479004, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8562015295028687, "num_tokens": 129524392.0, "step": 3393 }, { "epoch": 0.4317516855361913, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.731152534484863, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.842280924320221, "num_tokens": 129558452.0, "step": 3394 }, { "epoch": 0.43187889581478184, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.772306442260742, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.853752613067627, "num_tokens": 129597076.0, "step": 3395 }, { "epoch": 0.43200610609337237, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.723377227783203, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.849078893661499, "num_tokens": 129635861.0, "step": 3396 }, { "epoch": 0.43213331637196284, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.758428573608398, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8506864309310913, "num_tokens": 129673760.0, "step": 3397 }, { "epoch": 0.43226052665055337, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.74174976348877, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.857338547706604, "num_tokens": 129710730.0, "step": 3398 }, { "epoch": 0.4323877369291439, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.74133586883545, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8523643016815186, "num_tokens": 129748724.0, "step": 3399 }, { "epoch": 0.43251494720773437, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.767690658569336, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8688631057739258, "num_tokens": 129787478.0, "step": 3400 }, { "epoch": 0.4326421574863249, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.716751098632812, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8657775521278381, "num_tokens": 129829852.0, "step": 3401 }, { "epoch": 0.4327693677649154, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.76625919342041, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8547503352165222, "num_tokens": 129869279.0, "step": 3402 }, { "epoch": 0.4328965780435059, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.777261734008789, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8496637344360352, "num_tokens": 129903347.0, "step": 3403 }, { "epoch": 0.4330237883220964, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.805825233459473, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8652398586273193, "num_tokens": 129940028.0, "step": 3404 }, { "epoch": 0.43315099860068695, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.769519805908203, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8644729852676392, "num_tokens": 129976438.0, "step": 3405 }, { "epoch": 0.4332782088792774, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.706914901733398, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8608521223068237, "num_tokens": 130015698.0, "step": 3406 }, { "epoch": 0.43340541915786795, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.77493953704834, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.870347261428833, "num_tokens": 130056000.0, "step": 3407 }, { "epoch": 0.4335326294364585, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.728174209594727, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8535575866699219, "num_tokens": 130096990.0, "step": 3408 }, { "epoch": 0.43365983971504896, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.7550630569458, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8649120330810547, "num_tokens": 130128556.0, "step": 3409 }, { "epoch": 0.4337870499936395, "ewc_loss": 0.015625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.728501319885254, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8614449501037598, "num_tokens": 130170607.0, "step": 3410 }, { "epoch": 0.43391426027223, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.80197525024414, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8514494895935059, "num_tokens": 130212511.0, "step": 3411 }, { "epoch": 0.4340414705508205, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.810951232910156, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8490335941314697, "num_tokens": 130249246.0, "step": 3412 }, { "epoch": 0.434168680829411, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.724031448364258, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8759661912918091, "num_tokens": 130291250.0, "step": 3413 }, { "epoch": 0.43429589110800154, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.8655424118042, "learning_rate": 1e-06, "loss": 0.5028, "mean_token_accuracy": 0.8362668752670288, "num_tokens": 130327094.0, "step": 3414 }, { "epoch": 0.434423101386592, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.747912406921387, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8475964069366455, "num_tokens": 130366482.0, "step": 3415 }, { "epoch": 0.43455031166518254, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.73235034942627, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8557002544403076, "num_tokens": 130407416.0, "step": 3416 }, { "epoch": 0.43467752194377307, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.709190368652344, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8619180917739868, "num_tokens": 130447125.0, "step": 3417 }, { "epoch": 0.43480473222236354, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.785696029663086, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8624714612960815, "num_tokens": 130480955.0, "step": 3418 }, { "epoch": 0.43493194250095407, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.789022445678711, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8682633638381958, "num_tokens": 130511771.0, "step": 3419 }, { "epoch": 0.4350591527795446, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.708490371704102, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8550193309783936, "num_tokens": 130550080.0, "step": 3420 }, { "epoch": 0.4351863630581351, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.742108345031738, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.844041109085083, "num_tokens": 130588109.0, "step": 3421 }, { "epoch": 0.4353135733367256, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.785200119018555, "learning_rate": 1e-06, "loss": 0.5009, "mean_token_accuracy": 0.8422476053237915, "num_tokens": 130626832.0, "step": 3422 }, { "epoch": 0.43544078361531613, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.730239868164062, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8605196475982666, "num_tokens": 130662202.0, "step": 3423 }, { "epoch": 0.4355679938939066, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.755399703979492, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8517411351203918, "num_tokens": 130700747.0, "step": 3424 }, { "epoch": 0.43569520417249713, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.642117500305176, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8543832898139954, "num_tokens": 130742849.0, "step": 3425 }, { "epoch": 0.43582241445108766, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.758688926696777, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8694618940353394, "num_tokens": 130780156.0, "step": 3426 }, { "epoch": 0.43594962472967813, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.71880054473877, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8558851480484009, "num_tokens": 130825824.0, "step": 3427 }, { "epoch": 0.43607683500826866, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.733785629272461, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8646132349967957, "num_tokens": 130863603.0, "step": 3428 }, { "epoch": 0.4362040452868592, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.736043930053711, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8572903871536255, "num_tokens": 130906689.0, "step": 3429 }, { "epoch": 0.4363312555654497, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.776063919067383, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.853958785533905, "num_tokens": 130947785.0, "step": 3430 }, { "epoch": 0.4364584658440402, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.78072738647461, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.866676926612854, "num_tokens": 130985489.0, "step": 3431 }, { "epoch": 0.4365856761226307, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.687088966369629, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8650102615356445, "num_tokens": 131025274.0, "step": 3432 }, { "epoch": 0.43671288640122125, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.717201232910156, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8559486269950867, "num_tokens": 131060132.0, "step": 3433 }, { "epoch": 0.4368400966798117, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.71457576751709, "learning_rate": 1e-06, "loss": 0.4857, "mean_token_accuracy": 0.8453046679496765, "num_tokens": 131105109.0, "step": 3434 }, { "epoch": 0.43696730695840225, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.677358627319336, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8714576959609985, "num_tokens": 131142094.0, "step": 3435 }, { "epoch": 0.4370945172369928, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.744569778442383, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8551762104034424, "num_tokens": 131174648.0, "step": 3436 }, { "epoch": 0.43722172751558325, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.692052841186523, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8713780641555786, "num_tokens": 131213447.0, "step": 3437 }, { "epoch": 0.4373489377941738, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.795515060424805, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8523851633071899, "num_tokens": 131249622.0, "step": 3438 }, { "epoch": 0.4374761480727643, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.75588321685791, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8728140592575073, "num_tokens": 131285883.0, "step": 3439 }, { "epoch": 0.4376033583513548, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.751651763916016, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8597214818000793, "num_tokens": 131322774.0, "step": 3440 }, { "epoch": 0.4377305686299453, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.760415077209473, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8684383034706116, "num_tokens": 131364513.0, "step": 3441 }, { "epoch": 0.43785777890853583, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.744422912597656, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8467484712600708, "num_tokens": 131402299.0, "step": 3442 }, { "epoch": 0.4379849891871263, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.785889625549316, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8639086484909058, "num_tokens": 131447830.0, "step": 3443 }, { "epoch": 0.43811219946571683, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.818153381347656, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8499195575714111, "num_tokens": 131481058.0, "step": 3444 }, { "epoch": 0.43823940974430736, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.791905403137207, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8529559969902039, "num_tokens": 131518376.0, "step": 3445 }, { "epoch": 0.43836662002289783, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.81951904296875, "learning_rate": 1e-06, "loss": 0.5065, "mean_token_accuracy": 0.8420071005821228, "num_tokens": 131553702.0, "step": 3446 }, { "epoch": 0.43849383030148836, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.797874450683594, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8646115064620972, "num_tokens": 131587026.0, "step": 3447 }, { "epoch": 0.4386210405800789, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.775307655334473, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8626605272293091, "num_tokens": 131621145.0, "step": 3448 }, { "epoch": 0.43874825085866936, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.843405723571777, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8622406125068665, "num_tokens": 131661618.0, "step": 3449 }, { "epoch": 0.4388754611372599, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.786906242370605, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8507312536239624, "num_tokens": 131693840.0, "step": 3450 }, { "epoch": 0.4390026714158504, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.851064682006836, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8592320680618286, "num_tokens": 131730856.0, "step": 3451 }, { "epoch": 0.4391298816944409, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.737156867980957, "learning_rate": 1e-06, "loss": 0.5158, "mean_token_accuracy": 0.8333701491355896, "num_tokens": 131769172.0, "step": 3452 }, { "epoch": 0.4392570919730314, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.794419288635254, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8594772219657898, "num_tokens": 131810481.0, "step": 3453 }, { "epoch": 0.43938430225162195, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.804274559020996, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8587510585784912, "num_tokens": 131850742.0, "step": 3454 }, { "epoch": 0.4395115125302124, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.837858200073242, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8559088706970215, "num_tokens": 131886675.0, "step": 3455 }, { "epoch": 0.43963872280880295, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.831279754638672, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8625572919845581, "num_tokens": 131926831.0, "step": 3456 }, { "epoch": 0.4397659330873935, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.826653480529785, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8558571338653564, "num_tokens": 131962265.0, "step": 3457 }, { "epoch": 0.43989314336598395, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.86624813079834, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8526132702827454, "num_tokens": 131999906.0, "step": 3458 }, { "epoch": 0.4400203536445745, "ewc_loss": 0.015625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.804727554321289, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8734107613563538, "num_tokens": 132031330.0, "step": 3459 }, { "epoch": 0.440147563923165, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.841828346252441, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8554232120513916, "num_tokens": 132065049.0, "step": 3460 }, { "epoch": 0.4402747742017555, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.757940292358398, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8614088296890259, "num_tokens": 132100993.0, "step": 3461 }, { "epoch": 0.440401984480346, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.799722671508789, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8550148010253906, "num_tokens": 132138169.0, "step": 3462 }, { "epoch": 0.44052919475893654, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.848955154418945, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8641189336776733, "num_tokens": 132170498.0, "step": 3463 }, { "epoch": 0.440656405037527, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.76772403717041, "learning_rate": 1e-06, "loss": 0.5008, "mean_token_accuracy": 0.8378483653068542, "num_tokens": 132213583.0, "step": 3464 }, { "epoch": 0.44078361531611754, "ewc_loss": 0.015625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.8280611038208, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8524647355079651, "num_tokens": 132255427.0, "step": 3465 }, { "epoch": 0.44091082559470807, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.745644569396973, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8549697399139404, "num_tokens": 132295795.0, "step": 3466 }, { "epoch": 0.44103803587329854, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.80960750579834, "learning_rate": 1e-06, "loss": 0.4925, "mean_token_accuracy": 0.8391709327697754, "num_tokens": 132336881.0, "step": 3467 }, { "epoch": 0.44116524615188907, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.841558456420898, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8566812872886658, "num_tokens": 132375136.0, "step": 3468 }, { "epoch": 0.4412924564304796, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.806102752685547, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.853868842124939, "num_tokens": 132412893.0, "step": 3469 }, { "epoch": 0.44141966670907007, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.79719066619873, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8512000441551208, "num_tokens": 132444360.0, "step": 3470 }, { "epoch": 0.4415468769876606, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.758935928344727, "learning_rate": 1e-06, "loss": 0.5394, "mean_token_accuracy": 0.831233024597168, "num_tokens": 132485022.0, "step": 3471 }, { "epoch": 0.4416740872662511, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.819629669189453, "learning_rate": 1e-06, "loss": 0.5098, "mean_token_accuracy": 0.841670036315918, "num_tokens": 132522800.0, "step": 3472 }, { "epoch": 0.4418012975448416, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.733575820922852, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8597722053527832, "num_tokens": 132560527.0, "step": 3473 }, { "epoch": 0.4419285078234321, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.808743476867676, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.853793740272522, "num_tokens": 132601075.0, "step": 3474 }, { "epoch": 0.44205571810202265, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.795230865478516, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8478125929832458, "num_tokens": 132639851.0, "step": 3475 }, { "epoch": 0.4421829283806131, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.774469375610352, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8587479591369629, "num_tokens": 132679849.0, "step": 3476 }, { "epoch": 0.44231013865920366, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.772378921508789, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8617165088653564, "num_tokens": 132721020.0, "step": 3477 }, { "epoch": 0.4424373489377942, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.786285400390625, "learning_rate": 1e-06, "loss": 0.4756, "mean_token_accuracy": 0.8436989784240723, "num_tokens": 132760933.0, "step": 3478 }, { "epoch": 0.44256455921638466, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.758857727050781, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8675022125244141, "num_tokens": 132800627.0, "step": 3479 }, { "epoch": 0.4426917694949752, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.822126388549805, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8557385802268982, "num_tokens": 132833232.0, "step": 3480 }, { "epoch": 0.4428189797735657, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.759193420410156, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8547638654708862, "num_tokens": 132874670.0, "step": 3481 }, { "epoch": 0.44294619005215624, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.755484580993652, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.858600914478302, "num_tokens": 132916187.0, "step": 3482 }, { "epoch": 0.4430734003307467, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.759864807128906, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8590764999389648, "num_tokens": 132956948.0, "step": 3483 }, { "epoch": 0.44320061060933724, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.75757884979248, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8609727621078491, "num_tokens": 132995213.0, "step": 3484 }, { "epoch": 0.44332782088792777, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.818379402160645, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8536157011985779, "num_tokens": 133038850.0, "step": 3485 }, { "epoch": 0.44345503116651824, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.754353523254395, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.857975959777832, "num_tokens": 133080766.0, "step": 3486 }, { "epoch": 0.44358224144510877, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.805110931396484, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8534362316131592, "num_tokens": 133120845.0, "step": 3487 }, { "epoch": 0.4437094517236993, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.789496421813965, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8540222644805908, "num_tokens": 133166011.0, "step": 3488 }, { "epoch": 0.4438366620022898, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.78248405456543, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8548620343208313, "num_tokens": 133203507.0, "step": 3489 }, { "epoch": 0.4439638722808803, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.83941650390625, "learning_rate": 1e-06, "loss": 0.5146, "mean_token_accuracy": 0.8327186703681946, "num_tokens": 133240554.0, "step": 3490 }, { "epoch": 0.44409108255947083, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.816963195800781, "learning_rate": 1e-06, "loss": 0.5018, "mean_token_accuracy": 0.8396549820899963, "num_tokens": 133274987.0, "step": 3491 }, { "epoch": 0.4442182928380613, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.775134086608887, "learning_rate": 1e-06, "loss": 0.5255, "mean_token_accuracy": 0.8293452262878418, "num_tokens": 133319287.0, "step": 3492 }, { "epoch": 0.44434550311665183, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.792872428894043, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8606462478637695, "num_tokens": 133353670.0, "step": 3493 }, { "epoch": 0.44447271339524236, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.807912826538086, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8578863143920898, "num_tokens": 133391950.0, "step": 3494 }, { "epoch": 0.44459992367383283, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.774969100952148, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8611539602279663, "num_tokens": 133424758.0, "step": 3495 }, { "epoch": 0.44472713395242336, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.772356986999512, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8848089575767517, "num_tokens": 133459869.0, "step": 3496 }, { "epoch": 0.4448543442310139, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.83896255493164, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8547194004058838, "num_tokens": 133494598.0, "step": 3497 }, { "epoch": 0.44498155450960436, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.81641674041748, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8651736378669739, "num_tokens": 133536403.0, "step": 3498 }, { "epoch": 0.4451087647881949, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.866453170776367, "learning_rate": 1e-06, "loss": 0.5118, "mean_token_accuracy": 0.8354759216308594, "num_tokens": 133567319.0, "step": 3499 }, { "epoch": 0.4452359750667854, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.814457893371582, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.8494769334793091, "num_tokens": 133604829.0, "step": 3500 }, { "epoch": 0.4453631853453759, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.800809860229492, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8562114238739014, "num_tokens": 133638373.0, "step": 3501 }, { "epoch": 0.4454903956239664, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.812895774841309, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8677265644073486, "num_tokens": 133674110.0, "step": 3502 }, { "epoch": 0.44561760590255695, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.793208122253418, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8574791550636292, "num_tokens": 133711005.0, "step": 3503 }, { "epoch": 0.4457448161811474, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.740527153015137, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8689717054367065, "num_tokens": 133746258.0, "step": 3504 }, { "epoch": 0.44587202645973795, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.746050834655762, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8647986650466919, "num_tokens": 133790695.0, "step": 3505 }, { "epoch": 0.4459992367383285, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.757712364196777, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.853216290473938, "num_tokens": 133830181.0, "step": 3506 }, { "epoch": 0.44612644701691895, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.800053596496582, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8599058985710144, "num_tokens": 133866849.0, "step": 3507 }, { "epoch": 0.4462536572955095, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.805395126342773, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8560167551040649, "num_tokens": 133906812.0, "step": 3508 }, { "epoch": 0.4463808675741, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.690339088439941, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8495931625366211, "num_tokens": 133948956.0, "step": 3509 }, { "epoch": 0.4465080778526905, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.779736518859863, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.8392746448516846, "num_tokens": 133988764.0, "step": 3510 }, { "epoch": 0.446635288131281, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.769977569580078, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8613581657409668, "num_tokens": 134028148.0, "step": 3511 }, { "epoch": 0.44676249840987153, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.754725456237793, "learning_rate": 1e-06, "loss": 0.523, "mean_token_accuracy": 0.8322553038597107, "num_tokens": 134065909.0, "step": 3512 }, { "epoch": 0.446889708688462, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.755181312561035, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8629169464111328, "num_tokens": 134108342.0, "step": 3513 }, { "epoch": 0.44701691896705253, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.790841102600098, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8559024930000305, "num_tokens": 134145928.0, "step": 3514 }, { "epoch": 0.44714412924564306, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.792695999145508, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8524171710014343, "num_tokens": 134185955.0, "step": 3515 }, { "epoch": 0.44727133952423354, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.792051315307617, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8615573644638062, "num_tokens": 134225829.0, "step": 3516 }, { "epoch": 0.44739854980282406, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.861488342285156, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8468271493911743, "num_tokens": 134261381.0, "step": 3517 }, { "epoch": 0.4475257600814146, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.738759994506836, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8636544346809387, "num_tokens": 134294697.0, "step": 3518 }, { "epoch": 0.44765297036000506, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.79340934753418, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8769259452819824, "num_tokens": 134334349.0, "step": 3519 }, { "epoch": 0.4477801806385956, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.819804191589355, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8452235460281372, "num_tokens": 134373381.0, "step": 3520 }, { "epoch": 0.4479073909171861, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.773483276367188, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8614988923072815, "num_tokens": 134407952.0, "step": 3521 }, { "epoch": 0.4480346011957766, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.804974555969238, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8465477228164673, "num_tokens": 134458862.0, "step": 3522 }, { "epoch": 0.4481618114743671, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.854323387145996, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8526560068130493, "num_tokens": 134494007.0, "step": 3523 }, { "epoch": 0.44828902175295765, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.76111888885498, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8418149352073669, "num_tokens": 134534195.0, "step": 3524 }, { "epoch": 0.4484162320315481, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.794217109680176, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8649721145629883, "num_tokens": 134570241.0, "step": 3525 }, { "epoch": 0.44854344231013865, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.746639251708984, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8548089861869812, "num_tokens": 134611904.0, "step": 3526 }, { "epoch": 0.4486706525887292, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.761454582214355, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.84635329246521, "num_tokens": 134644426.0, "step": 3527 }, { "epoch": 0.44879786286731965, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.786480903625488, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8642831444740295, "num_tokens": 134678979.0, "step": 3528 }, { "epoch": 0.4489250731459102, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.763590812683105, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8590627312660217, "num_tokens": 134717270.0, "step": 3529 }, { "epoch": 0.4490522834245007, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.771709442138672, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8561825752258301, "num_tokens": 134755464.0, "step": 3530 }, { "epoch": 0.44917949370309124, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.765042304992676, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8482320308685303, "num_tokens": 134793737.0, "step": 3531 }, { "epoch": 0.4493067039816817, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.801206588745117, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8514602184295654, "num_tokens": 134834332.0, "step": 3532 }, { "epoch": 0.44943391426027224, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.741151809692383, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8668524622917175, "num_tokens": 134871027.0, "step": 3533 }, { "epoch": 0.44956112453886277, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.854707717895508, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8559134602546692, "num_tokens": 134908529.0, "step": 3534 }, { "epoch": 0.44968833481745324, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.766754150390625, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8647805452346802, "num_tokens": 134944177.0, "step": 3535 }, { "epoch": 0.44981554509604377, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.85193157196045, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8553143739700317, "num_tokens": 134977928.0, "step": 3536 }, { "epoch": 0.4499427553746343, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.825709342956543, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.857164740562439, "num_tokens": 135017554.0, "step": 3537 }, { "epoch": 0.45006996565322477, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.753066062927246, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8623912334442139, "num_tokens": 135058694.0, "step": 3538 }, { "epoch": 0.4501971759318153, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.832014083862305, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8583453893661499, "num_tokens": 135094037.0, "step": 3539 }, { "epoch": 0.4503243862104058, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.792071342468262, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8499013781547546, "num_tokens": 135132758.0, "step": 3540 }, { "epoch": 0.4504515964889963, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.854412078857422, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8690944314002991, "num_tokens": 135169398.0, "step": 3541 }, { "epoch": 0.4505788067675868, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.819994926452637, "learning_rate": 1e-06, "loss": 0.5141, "mean_token_accuracy": 0.8427060842514038, "num_tokens": 135204312.0, "step": 3542 }, { "epoch": 0.45070601704617735, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.847658157348633, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8530440926551819, "num_tokens": 135246287.0, "step": 3543 }, { "epoch": 0.4508332273247678, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.807368278503418, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8588877320289612, "num_tokens": 135280970.0, "step": 3544 }, { "epoch": 0.45096043760335836, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.830343246459961, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.878569483757019, "num_tokens": 135317873.0, "step": 3545 }, { "epoch": 0.4510876478819489, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.785869598388672, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8637804985046387, "num_tokens": 135354994.0, "step": 3546 }, { "epoch": 0.45121485816053936, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.784345626831055, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.851854145526886, "num_tokens": 135393050.0, "step": 3547 }, { "epoch": 0.4513420684391299, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.831608772277832, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8606508374214172, "num_tokens": 135431704.0, "step": 3548 }, { "epoch": 0.4514692787177204, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.797195434570312, "learning_rate": 1e-06, "loss": 0.5081, "mean_token_accuracy": 0.8341326713562012, "num_tokens": 135471306.0, "step": 3549 }, { "epoch": 0.4515964889963109, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.787944793701172, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8442344665527344, "num_tokens": 135509647.0, "step": 3550 }, { "epoch": 0.4517236992749014, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.83751392364502, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8640795946121216, "num_tokens": 135550469.0, "step": 3551 }, { "epoch": 0.45185090955349194, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.737582206726074, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8558446168899536, "num_tokens": 135589290.0, "step": 3552 }, { "epoch": 0.4519781198320824, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.830321311950684, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8507065773010254, "num_tokens": 135624416.0, "step": 3553 }, { "epoch": 0.45210533011067294, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.826746940612793, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8589845299720764, "num_tokens": 135662316.0, "step": 3554 }, { "epoch": 0.45223254038926347, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.837911605834961, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8584107160568237, "num_tokens": 135701739.0, "step": 3555 }, { "epoch": 0.45235975066785394, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.779186248779297, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8528915047645569, "num_tokens": 135741372.0, "step": 3556 }, { "epoch": 0.45248696094644447, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.783690452575684, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8783255815505981, "num_tokens": 135779945.0, "step": 3557 }, { "epoch": 0.452614171225035, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.865805625915527, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8514827489852905, "num_tokens": 135813179.0, "step": 3558 }, { "epoch": 0.4527413815036255, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.795402526855469, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.85921311378479, "num_tokens": 135855147.0, "step": 3559 }, { "epoch": 0.452868591782216, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.762214660644531, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8646558523178101, "num_tokens": 135892990.0, "step": 3560 }, { "epoch": 0.45299580206080653, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.772756576538086, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8751206398010254, "num_tokens": 135926911.0, "step": 3561 }, { "epoch": 0.453123012339397, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.791549682617188, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8655475378036499, "num_tokens": 135963530.0, "step": 3562 }, { "epoch": 0.45325022261798753, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.850146293640137, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8479578495025635, "num_tokens": 136000677.0, "step": 3563 }, { "epoch": 0.45337743289657806, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.786940574645996, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8464682102203369, "num_tokens": 136037283.0, "step": 3564 }, { "epoch": 0.45350464317516853, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.83462142944336, "learning_rate": 1e-06, "loss": 0.501, "mean_token_accuracy": 0.8393534421920776, "num_tokens": 136071454.0, "step": 3565 }, { "epoch": 0.45363185345375906, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.823145866394043, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8453906178474426, "num_tokens": 136105638.0, "step": 3566 }, { "epoch": 0.4537590637323496, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.802600860595703, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8605111837387085, "num_tokens": 136147968.0, "step": 3567 }, { "epoch": 0.45388627401094006, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.794973373413086, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.857610821723938, "num_tokens": 136190468.0, "step": 3568 }, { "epoch": 0.4540134842895306, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.806519508361816, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8609293699264526, "num_tokens": 136229363.0, "step": 3569 }, { "epoch": 0.4541406945681211, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.849828720092773, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8505664467811584, "num_tokens": 136268834.0, "step": 3570 }, { "epoch": 0.4542679048467116, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.800630569458008, "learning_rate": 1e-06, "loss": 0.5275, "mean_token_accuracy": 0.833797037601471, "num_tokens": 136306761.0, "step": 3571 }, { "epoch": 0.4543951151253021, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.853801727294922, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8558861613273621, "num_tokens": 136347736.0, "step": 3572 }, { "epoch": 0.45452232540389265, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.799967765808105, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8599608540534973, "num_tokens": 136381009.0, "step": 3573 }, { "epoch": 0.4546495356824831, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.840112686157227, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8677399754524231, "num_tokens": 136418050.0, "step": 3574 }, { "epoch": 0.45477674596107365, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.823387145996094, "learning_rate": 1e-06, "loss": 0.5005, "mean_token_accuracy": 0.838674008846283, "num_tokens": 136459109.0, "step": 3575 }, { "epoch": 0.4549039562396642, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.830504417419434, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8526164293289185, "num_tokens": 136497993.0, "step": 3576 }, { "epoch": 0.45503116651825465, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.867888450622559, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8586397171020508, "num_tokens": 136534446.0, "step": 3577 }, { "epoch": 0.4551583767968452, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.851117134094238, "learning_rate": 1e-06, "loss": 0.4951, "mean_token_accuracy": 0.8414026498794556, "num_tokens": 136571117.0, "step": 3578 }, { "epoch": 0.4552855870754357, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.814236640930176, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8638867139816284, "num_tokens": 136601706.0, "step": 3579 }, { "epoch": 0.4554127973540262, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.82201099395752, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8574201464653015, "num_tokens": 136639735.0, "step": 3580 }, { "epoch": 0.4555400076326167, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.82021427154541, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8496573567390442, "num_tokens": 136675647.0, "step": 3581 }, { "epoch": 0.45566721791120723, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.809245109558105, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8642131686210632, "num_tokens": 136713361.0, "step": 3582 }, { "epoch": 0.45579442818979776, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.827616691589355, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8557748794555664, "num_tokens": 136751440.0, "step": 3583 }, { "epoch": 0.45592163846838824, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.837859153747559, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.852400004863739, "num_tokens": 136786078.0, "step": 3584 }, { "epoch": 0.45604884874697876, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.787731170654297, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8636872172355652, "num_tokens": 136828313.0, "step": 3585 }, { "epoch": 0.4561760590255693, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.819623947143555, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8556793928146362, "num_tokens": 136862549.0, "step": 3586 }, { "epoch": 0.45630326930415976, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.78771686553955, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8593627214431763, "num_tokens": 136906423.0, "step": 3587 }, { "epoch": 0.4564304795827503, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.820340156555176, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8447344303131104, "num_tokens": 136941550.0, "step": 3588 }, { "epoch": 0.4565576898613408, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.702296257019043, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8593971729278564, "num_tokens": 136980234.0, "step": 3589 }, { "epoch": 0.4566849001399313, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.789135932922363, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8600200414657593, "num_tokens": 137023991.0, "step": 3590 }, { "epoch": 0.4568121104185218, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.77621078491211, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.8478708863258362, "num_tokens": 137063261.0, "step": 3591 }, { "epoch": 0.45693932069711235, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.853976249694824, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8659365177154541, "num_tokens": 137103935.0, "step": 3592 }, { "epoch": 0.4570665309757028, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.806631088256836, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8561832904815674, "num_tokens": 137140194.0, "step": 3593 }, { "epoch": 0.45719374125429335, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.799845695495605, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8471605181694031, "num_tokens": 137179086.0, "step": 3594 }, { "epoch": 0.4573209515328839, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.810815811157227, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8624392151832581, "num_tokens": 137216887.0, "step": 3595 }, { "epoch": 0.45744816181147435, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.765043258666992, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8720531463623047, "num_tokens": 137255029.0, "step": 3596 }, { "epoch": 0.4575753720900649, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.791059494018555, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8600542545318604, "num_tokens": 137291399.0, "step": 3597 }, { "epoch": 0.4577025823686554, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.844958305358887, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8609969615936279, "num_tokens": 137326569.0, "step": 3598 }, { "epoch": 0.4578297926472459, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.765767097473145, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8526362180709839, "num_tokens": 137370035.0, "step": 3599 }, { "epoch": 0.4579570029258364, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.828652381896973, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8527257442474365, "num_tokens": 137407969.0, "step": 3600 }, { "epoch": 0.45808421320442694, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.798114776611328, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8549846410751343, "num_tokens": 137447256.0, "step": 3601 }, { "epoch": 0.4582114234830174, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.799052238464355, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.8419880867004395, "num_tokens": 137479762.0, "step": 3602 }, { "epoch": 0.45833863376160794, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.796557426452637, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8500922918319702, "num_tokens": 137521823.0, "step": 3603 }, { "epoch": 0.45846584404019847, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.832829475402832, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8543621301651001, "num_tokens": 137558174.0, "step": 3604 }, { "epoch": 0.45859305431878894, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.81129264831543, "learning_rate": 1e-06, "loss": 0.5278, "mean_token_accuracy": 0.8330287933349609, "num_tokens": 137597265.0, "step": 3605 }, { "epoch": 0.45872026459737947, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.783220291137695, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8646863698959351, "num_tokens": 137635954.0, "step": 3606 }, { "epoch": 0.45884747487597, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.815643310546875, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8587892055511475, "num_tokens": 137676137.0, "step": 3607 }, { "epoch": 0.45897468515456047, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.841919898986816, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8595868349075317, "num_tokens": 137716723.0, "step": 3608 }, { "epoch": 0.459101895433151, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.77425479888916, "learning_rate": 1e-06, "loss": 0.4958, "mean_token_accuracy": 0.8382378816604614, "num_tokens": 137752435.0, "step": 3609 }, { "epoch": 0.4592291057117415, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.871461868286133, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8488600850105286, "num_tokens": 137786684.0, "step": 3610 }, { "epoch": 0.459356315990332, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.843753814697266, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8485824465751648, "num_tokens": 137830668.0, "step": 3611 }, { "epoch": 0.4594835262689225, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.792871475219727, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8627051711082458, "num_tokens": 137865262.0, "step": 3612 }, { "epoch": 0.45961073654751305, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.857645034790039, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8637713193893433, "num_tokens": 137898867.0, "step": 3613 }, { "epoch": 0.4597379468261035, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.84749698638916, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8584587574005127, "num_tokens": 137935108.0, "step": 3614 }, { "epoch": 0.45986515710469406, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.8126859664917, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8678415417671204, "num_tokens": 137977714.0, "step": 3615 }, { "epoch": 0.4599923673832846, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.884857177734375, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8684875965118408, "num_tokens": 138006345.0, "step": 3616 }, { "epoch": 0.46011957766187506, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.804793357849121, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8570960164070129, "num_tokens": 138045106.0, "step": 3617 }, { "epoch": 0.4602467879404656, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.900138854980469, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8521182537078857, "num_tokens": 138079993.0, "step": 3618 }, { "epoch": 0.4603739982190561, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.853861808776855, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8607618808746338, "num_tokens": 138118555.0, "step": 3619 }, { "epoch": 0.4605012084976466, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.807682037353516, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8642074465751648, "num_tokens": 138155096.0, "step": 3620 }, { "epoch": 0.4606284187762371, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.843451499938965, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8586010932922363, "num_tokens": 138194529.0, "step": 3621 }, { "epoch": 0.46075562905482764, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.866559982299805, "learning_rate": 1e-06, "loss": 0.5164, "mean_token_accuracy": 0.8314731121063232, "num_tokens": 138235520.0, "step": 3622 }, { "epoch": 0.4608828393334181, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.819621086120605, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8671389222145081, "num_tokens": 138271391.0, "step": 3623 }, { "epoch": 0.46101004961200864, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.80734634399414, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8514798879623413, "num_tokens": 138313923.0, "step": 3624 }, { "epoch": 0.46113725989059917, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.839852333068848, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8489882946014404, "num_tokens": 138355992.0, "step": 3625 }, { "epoch": 0.46126447016918964, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.85251235961914, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8768045902252197, "num_tokens": 138393074.0, "step": 3626 }, { "epoch": 0.4613916804477802, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.838858604431152, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8641068339347839, "num_tokens": 138427854.0, "step": 3627 }, { "epoch": 0.4615188907263707, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.851659774780273, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8665507435798645, "num_tokens": 138467359.0, "step": 3628 }, { "epoch": 0.4616461010049612, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.864336013793945, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8453587293624878, "num_tokens": 138505355.0, "step": 3629 }, { "epoch": 0.4617733112835517, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.856311798095703, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8608734607696533, "num_tokens": 138543703.0, "step": 3630 }, { "epoch": 0.46190052156214223, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.818296432495117, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8501908779144287, "num_tokens": 138579860.0, "step": 3631 }, { "epoch": 0.46202773184073276, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.858847618103027, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8672712445259094, "num_tokens": 138615755.0, "step": 3632 }, { "epoch": 0.46215494211932323, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.926431655883789, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.8521807193756104, "num_tokens": 138647607.0, "step": 3633 }, { "epoch": 0.46228215239791376, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.83782958984375, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8630237579345703, "num_tokens": 138679284.0, "step": 3634 }, { "epoch": 0.4624093626765043, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.857596397399902, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.866705060005188, "num_tokens": 138719690.0, "step": 3635 }, { "epoch": 0.46253657295509476, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.86351490020752, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8449628353118896, "num_tokens": 138758031.0, "step": 3636 }, { "epoch": 0.4626637832336853, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.846293449401855, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8570045232772827, "num_tokens": 138797283.0, "step": 3637 }, { "epoch": 0.4627909935122758, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.859209060668945, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8541334867477417, "num_tokens": 138834672.0, "step": 3638 }, { "epoch": 0.4629182037908663, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.783561706542969, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8687248229980469, "num_tokens": 138871105.0, "step": 3639 }, { "epoch": 0.4630454140694568, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.870172500610352, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8581064939498901, "num_tokens": 138903516.0, "step": 3640 }, { "epoch": 0.46317262434804735, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.811455726623535, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.855131983757019, "num_tokens": 138942021.0, "step": 3641 }, { "epoch": 0.4632998346266378, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.843416213989258, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8663697838783264, "num_tokens": 138978210.0, "step": 3642 }, { "epoch": 0.46342704490522835, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.79358196258545, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.859855592250824, "num_tokens": 139019122.0, "step": 3643 }, { "epoch": 0.4635542551838189, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.81663990020752, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8672173023223877, "num_tokens": 139062187.0, "step": 3644 }, { "epoch": 0.46368146546240935, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.8668794631958, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8766583800315857, "num_tokens": 139102960.0, "step": 3645 }, { "epoch": 0.4638086757409999, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.81535816192627, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8589766025543213, "num_tokens": 139137302.0, "step": 3646 }, { "epoch": 0.4639358860195904, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.800529479980469, "learning_rate": 1e-06, "loss": 0.5074, "mean_token_accuracy": 0.838672935962677, "num_tokens": 139177195.0, "step": 3647 }, { "epoch": 0.4640630962981809, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.829190254211426, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8542282581329346, "num_tokens": 139214722.0, "step": 3648 }, { "epoch": 0.4641903065767714, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.858985900878906, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8559553027153015, "num_tokens": 139248569.0, "step": 3649 }, { "epoch": 0.46431751685536193, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.832422256469727, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8639248609542847, "num_tokens": 139289141.0, "step": 3650 }, { "epoch": 0.4644447271339524, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.821708679199219, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8750844597816467, "num_tokens": 139330791.0, "step": 3651 }, { "epoch": 0.46457193741254293, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.845404624938965, "learning_rate": 1e-06, "loss": 0.5384, "mean_token_accuracy": 0.8262870907783508, "num_tokens": 139373009.0, "step": 3652 }, { "epoch": 0.46469914769113346, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.838064193725586, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8558850884437561, "num_tokens": 139408894.0, "step": 3653 }, { "epoch": 0.46482635796972394, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.820924758911133, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8559184074401855, "num_tokens": 139444040.0, "step": 3654 }, { "epoch": 0.46495356824831446, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.854740142822266, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8636345267295837, "num_tokens": 139479675.0, "step": 3655 }, { "epoch": 0.465080778526905, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.813652992248535, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8641073107719421, "num_tokens": 139521792.0, "step": 3656 }, { "epoch": 0.46520798880549546, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.872330665588379, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8637113571166992, "num_tokens": 139563471.0, "step": 3657 }, { "epoch": 0.465335199084086, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.810882568359375, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8580462336540222, "num_tokens": 139604702.0, "step": 3658 }, { "epoch": 0.4654624093626765, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.842222213745117, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8547764420509338, "num_tokens": 139644573.0, "step": 3659 }, { "epoch": 0.465589619641267, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.882394790649414, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8432379364967346, "num_tokens": 139680924.0, "step": 3660 }, { "epoch": 0.4657168299198575, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.787477493286133, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8500394225120544, "num_tokens": 139724369.0, "step": 3661 }, { "epoch": 0.46584404019844805, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.837974548339844, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8580577373504639, "num_tokens": 139769412.0, "step": 3662 }, { "epoch": 0.4659712504770385, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.861668586730957, "learning_rate": 1e-06, "loss": 0.5125, "mean_token_accuracy": 0.8337441086769104, "num_tokens": 139808066.0, "step": 3663 }, { "epoch": 0.46609846075562905, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.84030532836914, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8628779649734497, "num_tokens": 139851169.0, "step": 3664 }, { "epoch": 0.4662256710342196, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.855868339538574, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8676047325134277, "num_tokens": 139890550.0, "step": 3665 }, { "epoch": 0.46635288131281005, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.841352462768555, "learning_rate": 1e-06, "loss": 0.5222, "mean_token_accuracy": 0.8314206600189209, "num_tokens": 139931496.0, "step": 3666 }, { "epoch": 0.4664800915914006, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.848538398742676, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8479806184768677, "num_tokens": 139965872.0, "step": 3667 }, { "epoch": 0.4666073018699911, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.89294147491455, "learning_rate": 1e-06, "loss": 0.5118, "mean_token_accuracy": 0.8425961136817932, "num_tokens": 140000394.0, "step": 3668 }, { "epoch": 0.4667345121485816, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.81997013092041, "learning_rate": 1e-06, "loss": 0.4914, "mean_token_accuracy": 0.8397092223167419, "num_tokens": 140039967.0, "step": 3669 }, { "epoch": 0.4668617224271721, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.874077796936035, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8580223321914673, "num_tokens": 140075739.0, "step": 3670 }, { "epoch": 0.46698893270576264, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.897923469543457, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8623594045639038, "num_tokens": 140110016.0, "step": 3671 }, { "epoch": 0.4671161429843531, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.824800491333008, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8477344512939453, "num_tokens": 140147905.0, "step": 3672 }, { "epoch": 0.46724335326294364, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.938395500183105, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8513222932815552, "num_tokens": 140182284.0, "step": 3673 }, { "epoch": 0.46737056354153417, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.800780296325684, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8496003150939941, "num_tokens": 140220669.0, "step": 3674 }, { "epoch": 0.46749777382012464, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.844882011413574, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8428573608398438, "num_tokens": 140258399.0, "step": 3675 }, { "epoch": 0.46762498409871517, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.851080894470215, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8705103397369385, "num_tokens": 140296015.0, "step": 3676 }, { "epoch": 0.4677521943773057, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.85237979888916, "learning_rate": 1e-06, "loss": 0.4783, "mean_token_accuracy": 0.8461675047874451, "num_tokens": 140338618.0, "step": 3677 }, { "epoch": 0.46787940465589617, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.840654373168945, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8612747192382812, "num_tokens": 140379151.0, "step": 3678 }, { "epoch": 0.4680066149344867, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.819990158081055, "learning_rate": 1e-06, "loss": 0.4846, "mean_token_accuracy": 0.8428758382797241, "num_tokens": 140417740.0, "step": 3679 }, { "epoch": 0.4681338252130772, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.87536907196045, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8540139198303223, "num_tokens": 140452454.0, "step": 3680 }, { "epoch": 0.46826103549166775, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.842822074890137, "learning_rate": 1e-06, "loss": 0.5219, "mean_token_accuracy": 0.8332074880599976, "num_tokens": 140492981.0, "step": 3681 }, { "epoch": 0.4683882457702582, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.82178783416748, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8457657098770142, "num_tokens": 140534131.0, "step": 3682 }, { "epoch": 0.46851545604884876, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.84575080871582, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.8490276336669922, "num_tokens": 140576761.0, "step": 3683 }, { "epoch": 0.4686426663274393, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.984624862670898, "learning_rate": 1e-06, "loss": 0.5301, "mean_token_accuracy": 0.8367553353309631, "num_tokens": 140609538.0, "step": 3684 }, { "epoch": 0.46876987660602976, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.887214660644531, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8520691990852356, "num_tokens": 140641802.0, "step": 3685 }, { "epoch": 0.4688970868846203, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.824711799621582, "learning_rate": 1e-06, "loss": 0.486, "mean_token_accuracy": 0.8442636132240295, "num_tokens": 140685832.0, "step": 3686 }, { "epoch": 0.4690242971632108, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.881156921386719, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8436868190765381, "num_tokens": 140725333.0, "step": 3687 }, { "epoch": 0.4691515074418013, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.854485511779785, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8540894985198975, "num_tokens": 140760838.0, "step": 3688 }, { "epoch": 0.4692787177203918, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.907184600830078, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8453238010406494, "num_tokens": 140797367.0, "step": 3689 }, { "epoch": 0.46940592799898234, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.791913986206055, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8534814715385437, "num_tokens": 140839778.0, "step": 3690 }, { "epoch": 0.4695331382775728, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.887421607971191, "learning_rate": 1e-06, "loss": 0.4649, "mean_token_accuracy": 0.8508291244506836, "num_tokens": 140880340.0, "step": 3691 }, { "epoch": 0.46966034855616334, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.815767288208008, "learning_rate": 1e-06, "loss": 0.4771, "mean_token_accuracy": 0.8497000336647034, "num_tokens": 140915451.0, "step": 3692 }, { "epoch": 0.46978755883475387, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.870026588439941, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8527256846427917, "num_tokens": 140956967.0, "step": 3693 }, { "epoch": 0.46991476911334434, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.833513259887695, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8707005977630615, "num_tokens": 140996640.0, "step": 3694 }, { "epoch": 0.47004197939193487, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.829809188842773, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.854572057723999, "num_tokens": 141036094.0, "step": 3695 }, { "epoch": 0.4701691896705254, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.940214157104492, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8456065654754639, "num_tokens": 141069033.0, "step": 3696 }, { "epoch": 0.4702963999491159, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.848087310791016, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8722753524780273, "num_tokens": 141111075.0, "step": 3697 }, { "epoch": 0.4704236102277064, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.916748046875, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8714622855186462, "num_tokens": 141145270.0, "step": 3698 }, { "epoch": 0.47055082050629693, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.867806434631348, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8700636029243469, "num_tokens": 141182132.0, "step": 3699 }, { "epoch": 0.4706780307848874, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.898587226867676, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8563234806060791, "num_tokens": 141217455.0, "step": 3700 }, { "epoch": 0.47080524106347793, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.880160331726074, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8549680709838867, "num_tokens": 141249680.0, "step": 3701 }, { "epoch": 0.47093245134206846, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.913326263427734, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8560739159584045, "num_tokens": 141292085.0, "step": 3702 }, { "epoch": 0.47105966162065893, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.824576377868652, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8647640943527222, "num_tokens": 141331455.0, "step": 3703 }, { "epoch": 0.47118687189924946, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.965858459472656, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8561788201332092, "num_tokens": 141365623.0, "step": 3704 }, { "epoch": 0.47131408217784, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.861543655395508, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8473978042602539, "num_tokens": 141399033.0, "step": 3705 }, { "epoch": 0.47144129245643046, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.947514533996582, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8629188537597656, "num_tokens": 141430436.0, "step": 3706 }, { "epoch": 0.471568502735021, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.880025863647461, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8729996681213379, "num_tokens": 141463169.0, "step": 3707 }, { "epoch": 0.4716957130136115, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.919638633728027, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8740724325180054, "num_tokens": 141495524.0, "step": 3708 }, { "epoch": 0.471822923292202, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.824477195739746, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8680967688560486, "num_tokens": 141533099.0, "step": 3709 }, { "epoch": 0.4719501335707925, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.9247407913208, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8610657453536987, "num_tokens": 141576389.0, "step": 3710 }, { "epoch": 0.47207734384938305, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.889412879943848, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8590118288993835, "num_tokens": 141611870.0, "step": 3711 }, { "epoch": 0.4722045541279735, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.922988891601562, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8684767484664917, "num_tokens": 141648539.0, "step": 3712 }, { "epoch": 0.47233176440656405, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.864093780517578, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8605331182479858, "num_tokens": 141688549.0, "step": 3713 }, { "epoch": 0.4724589746851546, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.90650749206543, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8657699823379517, "num_tokens": 141725294.0, "step": 3714 }, { "epoch": 0.47258618496374505, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.882441520690918, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8412265181541443, "num_tokens": 141756761.0, "step": 3715 }, { "epoch": 0.4727133952423356, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.967734336853027, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.8416948318481445, "num_tokens": 141790417.0, "step": 3716 }, { "epoch": 0.4728406055209261, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.84045124053955, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8684926629066467, "num_tokens": 141830749.0, "step": 3717 }, { "epoch": 0.4729678157995166, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.918097496032715, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8618035316467285, "num_tokens": 141868433.0, "step": 3718 }, { "epoch": 0.4730950260781071, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.864368438720703, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8749881386756897, "num_tokens": 141907140.0, "step": 3719 }, { "epoch": 0.47322223635669763, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.875481605529785, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8709607720375061, "num_tokens": 141947436.0, "step": 3720 }, { "epoch": 0.4733494466352881, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.85952091217041, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8505067825317383, "num_tokens": 141986876.0, "step": 3721 }, { "epoch": 0.47347665691387864, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.875404357910156, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8609076738357544, "num_tokens": 142025806.0, "step": 3722 }, { "epoch": 0.47360386719246916, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.86854076385498, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.857286810874939, "num_tokens": 142057932.0, "step": 3723 }, { "epoch": 0.47373107747105964, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.92474365234375, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8607289791107178, "num_tokens": 142088880.0, "step": 3724 }, { "epoch": 0.47385828774965016, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 10.480257987976074, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8702008724212646, "num_tokens": 142124959.0, "step": 3725 }, { "epoch": 0.4739854980282407, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.862537384033203, "learning_rate": 1e-06, "loss": 0.5051, "mean_token_accuracy": 0.8433632850646973, "num_tokens": 142160667.0, "step": 3726 }, { "epoch": 0.47411270830683117, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.74142074584961, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8567516803741455, "num_tokens": 142205017.0, "step": 3727 }, { "epoch": 0.4742399185854217, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.856782913208008, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8641789555549622, "num_tokens": 142241479.0, "step": 3728 }, { "epoch": 0.4743671288640122, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.877115249633789, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.858089804649353, "num_tokens": 142280881.0, "step": 3729 }, { "epoch": 0.4744943391426027, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.837554931640625, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8649895191192627, "num_tokens": 142328248.0, "step": 3730 }, { "epoch": 0.4746215494211932, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.920985221862793, "learning_rate": 1e-06, "loss": 0.5058, "mean_token_accuracy": 0.8416378498077393, "num_tokens": 142364872.0, "step": 3731 }, { "epoch": 0.47474875969978375, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.851975440979004, "learning_rate": 1e-06, "loss": 0.4975, "mean_token_accuracy": 0.8400716781616211, "num_tokens": 142400494.0, "step": 3732 }, { "epoch": 0.4748759699783743, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.91202449798584, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8555418848991394, "num_tokens": 142433739.0, "step": 3733 }, { "epoch": 0.47500318025696475, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.854682922363281, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.866456151008606, "num_tokens": 142469777.0, "step": 3734 }, { "epoch": 0.4751303905355553, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.840706825256348, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8479399681091309, "num_tokens": 142510052.0, "step": 3735 }, { "epoch": 0.4752576008141458, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.920199394226074, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.845782995223999, "num_tokens": 142545689.0, "step": 3736 }, { "epoch": 0.4753848110927363, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.882135391235352, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8569263219833374, "num_tokens": 142585916.0, "step": 3737 }, { "epoch": 0.4755120213713268, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.883118629455566, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8646979331970215, "num_tokens": 142623019.0, "step": 3738 }, { "epoch": 0.47563923164991734, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.915602684020996, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8538873195648193, "num_tokens": 142660986.0, "step": 3739 }, { "epoch": 0.4757664419285078, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.870126724243164, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8614522814750671, "num_tokens": 142702876.0, "step": 3740 }, { "epoch": 0.47589365220709834, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.937146186828613, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8458282351493835, "num_tokens": 142744082.0, "step": 3741 }, { "epoch": 0.47602086248568887, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.936522483825684, "learning_rate": 1e-06, "loss": 0.4938, "mean_token_accuracy": 0.8391502499580383, "num_tokens": 142777511.0, "step": 3742 }, { "epoch": 0.47614807276427934, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.88093376159668, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8519087433815002, "num_tokens": 142818882.0, "step": 3743 }, { "epoch": 0.47627528304286987, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.984153747558594, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8605628609657288, "num_tokens": 142852771.0, "step": 3744 }, { "epoch": 0.4764024933214604, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.860795974731445, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8573753833770752, "num_tokens": 142888398.0, "step": 3745 }, { "epoch": 0.47652970360005087, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.896921157836914, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8730332851409912, "num_tokens": 142924789.0, "step": 3746 }, { "epoch": 0.4766569138786414, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.892097473144531, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.865538477897644, "num_tokens": 142958843.0, "step": 3747 }, { "epoch": 0.4767841241572319, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.914347648620605, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8687050938606262, "num_tokens": 142996408.0, "step": 3748 }, { "epoch": 0.4769113344358224, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.869831085205078, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8501231670379639, "num_tokens": 143040031.0, "step": 3749 }, { "epoch": 0.4770385447144129, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.865554809570312, "learning_rate": 1e-06, "loss": 0.5093, "mean_token_accuracy": 0.8344476222991943, "num_tokens": 143081648.0, "step": 3750 }, { "epoch": 0.47716575499300345, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.908940315246582, "learning_rate": 1e-06, "loss": 0.5037, "mean_token_accuracy": 0.8368046283721924, "num_tokens": 143118913.0, "step": 3751 }, { "epoch": 0.4772929652715939, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.929197311401367, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8659158945083618, "num_tokens": 143153858.0, "step": 3752 }, { "epoch": 0.47742017555018446, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.897079467773438, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8682088255882263, "num_tokens": 143190977.0, "step": 3753 }, { "epoch": 0.477547385828775, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.905394554138184, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8492233753204346, "num_tokens": 143228009.0, "step": 3754 }, { "epoch": 0.47767459610736546, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.948064804077148, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8470200896263123, "num_tokens": 143269820.0, "step": 3755 }, { "epoch": 0.477801806385956, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.934639930725098, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.861501157283783, "num_tokens": 143303441.0, "step": 3756 }, { "epoch": 0.4779290166645465, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.934985160827637, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8489974141120911, "num_tokens": 143338748.0, "step": 3757 }, { "epoch": 0.478056226943137, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.910977363586426, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8581275939941406, "num_tokens": 143373900.0, "step": 3758 }, { "epoch": 0.4781834372217275, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.933881759643555, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8523167371749878, "num_tokens": 143411190.0, "step": 3759 }, { "epoch": 0.47831064750031804, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.957218170166016, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8526865243911743, "num_tokens": 143449470.0, "step": 3760 }, { "epoch": 0.4784378577789085, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.910357475280762, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8626416921615601, "num_tokens": 143485868.0, "step": 3761 }, { "epoch": 0.47856506805749904, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.939472198486328, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.865939736366272, "num_tokens": 143520334.0, "step": 3762 }, { "epoch": 0.47869227833608957, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.876218795776367, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8504701852798462, "num_tokens": 143557752.0, "step": 3763 }, { "epoch": 0.47881948861468004, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.951096534729004, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8565368056297302, "num_tokens": 143593385.0, "step": 3764 }, { "epoch": 0.4789466988932706, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.92908763885498, "learning_rate": 1e-06, "loss": 0.5083, "mean_token_accuracy": 0.837084949016571, "num_tokens": 143631975.0, "step": 3765 }, { "epoch": 0.4790739091718611, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.93466854095459, "learning_rate": 1e-06, "loss": 0.4933, "mean_token_accuracy": 0.8392380475997925, "num_tokens": 143668382.0, "step": 3766 }, { "epoch": 0.4792011194504516, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.93214225769043, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8425871133804321, "num_tokens": 143706623.0, "step": 3767 }, { "epoch": 0.4793283297290421, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.892743110656738, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8610983490943909, "num_tokens": 143745664.0, "step": 3768 }, { "epoch": 0.47945554000763263, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.97329330444336, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8481937646865845, "num_tokens": 143780761.0, "step": 3769 }, { "epoch": 0.4795827502862231, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.872750282287598, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8618648052215576, "num_tokens": 143825298.0, "step": 3770 }, { "epoch": 0.47970996056481363, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.914250373840332, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8768386840820312, "num_tokens": 143859855.0, "step": 3771 }, { "epoch": 0.47983717084340416, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.94156265258789, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8701249361038208, "num_tokens": 143898553.0, "step": 3772 }, { "epoch": 0.47996438112199463, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 10.008950233459473, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8543751239776611, "num_tokens": 143935362.0, "step": 3773 }, { "epoch": 0.48009159140058516, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.882341384887695, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8574944734573364, "num_tokens": 143978692.0, "step": 3774 }, { "epoch": 0.4802188016791757, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.933791160583496, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8647891283035278, "num_tokens": 144020951.0, "step": 3775 }, { "epoch": 0.48034601195776616, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.933663368225098, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8660048246383667, "num_tokens": 144058341.0, "step": 3776 }, { "epoch": 0.4804732222363567, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.959961891174316, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8494701385498047, "num_tokens": 144098232.0, "step": 3777 }, { "epoch": 0.4806004325149472, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.959189414978027, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8513311743736267, "num_tokens": 144137537.0, "step": 3778 }, { "epoch": 0.4807276427935377, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.916661262512207, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8620063066482544, "num_tokens": 144178459.0, "step": 3779 }, { "epoch": 0.4808548530721282, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.910948753356934, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8522465229034424, "num_tokens": 144219257.0, "step": 3780 }, { "epoch": 0.48098206335071875, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.91933822631836, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.863376259803772, "num_tokens": 144262232.0, "step": 3781 }, { "epoch": 0.4811092736293093, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.969213485717773, "learning_rate": 1e-06, "loss": 0.5019, "mean_token_accuracy": 0.8352872133255005, "num_tokens": 144297144.0, "step": 3782 }, { "epoch": 0.48123648390789975, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.883594512939453, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8575056791305542, "num_tokens": 144336078.0, "step": 3783 }, { "epoch": 0.4813636941864903, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.88544750213623, "learning_rate": 1e-06, "loss": 0.4728, "mean_token_accuracy": 0.8485790491104126, "num_tokens": 144380162.0, "step": 3784 }, { "epoch": 0.4814909044650808, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.960131645202637, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.8426094055175781, "num_tokens": 144416735.0, "step": 3785 }, { "epoch": 0.4816181147436713, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.835315704345703, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8460537791252136, "num_tokens": 144452780.0, "step": 3786 }, { "epoch": 0.4817453250222618, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.94056510925293, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8629773855209351, "num_tokens": 144494155.0, "step": 3787 }, { "epoch": 0.48187253530085233, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.861658096313477, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8663640022277832, "num_tokens": 144533989.0, "step": 3788 }, { "epoch": 0.4819997455794428, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.90678882598877, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8528887629508972, "num_tokens": 144577409.0, "step": 3789 }, { "epoch": 0.48212695585803333, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.963181495666504, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.849443793296814, "num_tokens": 144618501.0, "step": 3790 }, { "epoch": 0.48225416613662386, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.874755859375, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8635894060134888, "num_tokens": 144659454.0, "step": 3791 }, { "epoch": 0.48238137641521434, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.906242370605469, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8623492121696472, "num_tokens": 144695580.0, "step": 3792 }, { "epoch": 0.48250858669380486, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.919559478759766, "learning_rate": 1e-06, "loss": 0.4998, "mean_token_accuracy": 0.8401969075202942, "num_tokens": 144735586.0, "step": 3793 }, { "epoch": 0.4826357969723954, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.855996131896973, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8583263158798218, "num_tokens": 144776371.0, "step": 3794 }, { "epoch": 0.48276300725098586, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.893061637878418, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8512617349624634, "num_tokens": 144816724.0, "step": 3795 }, { "epoch": 0.4828902175295764, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.99390697479248, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8634124994277954, "num_tokens": 144854591.0, "step": 3796 }, { "epoch": 0.4830174278081669, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.874984741210938, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8580683469772339, "num_tokens": 144893694.0, "step": 3797 }, { "epoch": 0.4831446380867574, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.932587623596191, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8539133071899414, "num_tokens": 144932985.0, "step": 3798 }, { "epoch": 0.4832718483653479, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.938831329345703, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8568107485771179, "num_tokens": 144967823.0, "step": 3799 }, { "epoch": 0.48339905864393845, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.916495323181152, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8771060109138489, "num_tokens": 145007750.0, "step": 3800 }, { "epoch": 0.4835262689225289, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.958003997802734, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8513680696487427, "num_tokens": 145048596.0, "step": 3801 }, { "epoch": 0.48365347920111945, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.956652641296387, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8466259241104126, "num_tokens": 145080868.0, "step": 3802 }, { "epoch": 0.48378068947971, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.977397918701172, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8665859699249268, "num_tokens": 145115865.0, "step": 3803 }, { "epoch": 0.48390789975830045, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.949910163879395, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8695467114448547, "num_tokens": 145151748.0, "step": 3804 }, { "epoch": 0.484035110036891, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.969038009643555, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8570182919502258, "num_tokens": 145187074.0, "step": 3805 }, { "epoch": 0.4841623203154815, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.917476654052734, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8747327923774719, "num_tokens": 145224855.0, "step": 3806 }, { "epoch": 0.484289530594072, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.899852752685547, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8658967018127441, "num_tokens": 145266538.0, "step": 3807 }, { "epoch": 0.4844167408726625, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.91091537475586, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8619387149810791, "num_tokens": 145303560.0, "step": 3808 }, { "epoch": 0.48454395115125304, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.9139986038208, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8587585687637329, "num_tokens": 145345985.0, "step": 3809 }, { "epoch": 0.4846711614298435, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.987687110900879, "learning_rate": 1e-06, "loss": 0.4806, "mean_token_accuracy": 0.8465368151664734, "num_tokens": 145382132.0, "step": 3810 }, { "epoch": 0.48479837170843404, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.94365406036377, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8478838205337524, "num_tokens": 145418992.0, "step": 3811 }, { "epoch": 0.48492558198702457, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.934107780456543, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8559256792068481, "num_tokens": 145455892.0, "step": 3812 }, { "epoch": 0.48505279226561504, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.906306266784668, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.855694055557251, "num_tokens": 145497217.0, "step": 3813 }, { "epoch": 0.48518000254420557, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.928105354309082, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8633792400360107, "num_tokens": 145533822.0, "step": 3814 }, { "epoch": 0.4853072128227961, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.943593978881836, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8770086765289307, "num_tokens": 145572967.0, "step": 3815 }, { "epoch": 0.48543442310138657, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.988273620605469, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8718709945678711, "num_tokens": 145609436.0, "step": 3816 }, { "epoch": 0.4855616333799771, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.971571922302246, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8692625761032104, "num_tokens": 145645083.0, "step": 3817 }, { "epoch": 0.4856888436585676, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.96833324432373, "learning_rate": 1e-06, "loss": 0.5202, "mean_token_accuracy": 0.8389871716499329, "num_tokens": 145680249.0, "step": 3818 }, { "epoch": 0.4858160539371581, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.99565601348877, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8626703023910522, "num_tokens": 145718569.0, "step": 3819 }, { "epoch": 0.4859432642157486, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.980873107910156, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8416690230369568, "num_tokens": 145751868.0, "step": 3820 }, { "epoch": 0.48607047449433916, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.983386993408203, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.862358033657074, "num_tokens": 145787881.0, "step": 3821 }, { "epoch": 0.48619768477292963, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.948843955993652, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8643618226051331, "num_tokens": 145824010.0, "step": 3822 }, { "epoch": 0.48632489505152016, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.943960189819336, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8640154004096985, "num_tokens": 145860981.0, "step": 3823 }, { "epoch": 0.4864521053301107, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.973111152648926, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8623380064964294, "num_tokens": 145900687.0, "step": 3824 }, { "epoch": 0.48657931560870116, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.908977508544922, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8495023846626282, "num_tokens": 145938888.0, "step": 3825 }, { "epoch": 0.4867065258872917, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.937346458435059, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8539630174636841, "num_tokens": 145987060.0, "step": 3826 }, { "epoch": 0.4868337361658822, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.931462287902832, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8530409336090088, "num_tokens": 146026447.0, "step": 3827 }, { "epoch": 0.4869609464444727, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.937077522277832, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8596669435501099, "num_tokens": 146067099.0, "step": 3828 }, { "epoch": 0.4870881567230632, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.93264102935791, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.844805896282196, "num_tokens": 146110767.0, "step": 3829 }, { "epoch": 0.48721536700165374, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.984384536743164, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8498834371566772, "num_tokens": 146147603.0, "step": 3830 }, { "epoch": 0.48734257728024427, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.953425407409668, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8680625557899475, "num_tokens": 146188821.0, "step": 3831 }, { "epoch": 0.48746978755883474, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.935507774353027, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8547898530960083, "num_tokens": 146231886.0, "step": 3832 }, { "epoch": 0.48759699783742527, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.947327613830566, "learning_rate": 1e-06, "loss": 0.5037, "mean_token_accuracy": 0.8359096050262451, "num_tokens": 146271965.0, "step": 3833 }, { "epoch": 0.4877242081160158, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.969488143920898, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.844427764415741, "num_tokens": 146316686.0, "step": 3834 }, { "epoch": 0.4878514183946063, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.899868965148926, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8473395109176636, "num_tokens": 146355098.0, "step": 3835 }, { "epoch": 0.4879786286731968, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.944307327270508, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8632323741912842, "num_tokens": 146390666.0, "step": 3836 }, { "epoch": 0.48810583895178733, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.891204833984375, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8643858432769775, "num_tokens": 146426740.0, "step": 3837 }, { "epoch": 0.4882330492303778, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.9630126953125, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8479769229888916, "num_tokens": 146462668.0, "step": 3838 }, { "epoch": 0.48836025950896833, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.956891059875488, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8499497771263123, "num_tokens": 146500160.0, "step": 3839 }, { "epoch": 0.48848746978755886, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.859310150146484, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8525748252868652, "num_tokens": 146542963.0, "step": 3840 }, { "epoch": 0.48861468006614933, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.918766021728516, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8647792339324951, "num_tokens": 146580649.0, "step": 3841 }, { "epoch": 0.48874189034473986, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.923933982849121, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8643084764480591, "num_tokens": 146620823.0, "step": 3842 }, { "epoch": 0.4888691006233304, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.915193557739258, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8677629232406616, "num_tokens": 146663700.0, "step": 3843 }, { "epoch": 0.48899631090192086, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.93665885925293, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8526129722595215, "num_tokens": 146698211.0, "step": 3844 }, { "epoch": 0.4891235211805114, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.900525093078613, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8556381464004517, "num_tokens": 146738380.0, "step": 3845 }, { "epoch": 0.4892507314591019, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.904714584350586, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8602771162986755, "num_tokens": 146775413.0, "step": 3846 }, { "epoch": 0.4893779417376924, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.933786392211914, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8599361181259155, "num_tokens": 146814774.0, "step": 3847 }, { "epoch": 0.4895051520162829, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.958045959472656, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8671259880065918, "num_tokens": 146853957.0, "step": 3848 }, { "epoch": 0.48963236229487345, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.955759048461914, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8503150939941406, "num_tokens": 146889060.0, "step": 3849 }, { "epoch": 0.4897595725734639, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.937023162841797, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8672610521316528, "num_tokens": 146929849.0, "step": 3850 }, { "epoch": 0.48988678285205445, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.960773468017578, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8596660494804382, "num_tokens": 146968376.0, "step": 3851 }, { "epoch": 0.490013993130645, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.948641777038574, "learning_rate": 1e-06, "loss": 0.4984, "mean_token_accuracy": 0.8415914177894592, "num_tokens": 147002871.0, "step": 3852 }, { "epoch": 0.49014120340923545, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.977995872497559, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8540806770324707, "num_tokens": 147039412.0, "step": 3853 }, { "epoch": 0.490268413687826, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.918573379516602, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8614845275878906, "num_tokens": 147080307.0, "step": 3854 }, { "epoch": 0.4903956239664165, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 10.045621871948242, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8654353618621826, "num_tokens": 147120131.0, "step": 3855 }, { "epoch": 0.490522834245007, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.891000747680664, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8512594103813171, "num_tokens": 147155760.0, "step": 3856 }, { "epoch": 0.4906500445235975, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.959016799926758, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8621534109115601, "num_tokens": 147195406.0, "step": 3857 }, { "epoch": 0.49077725480218803, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.95638656616211, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8711051940917969, "num_tokens": 147235688.0, "step": 3858 }, { "epoch": 0.4909044650807785, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.945537567138672, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8659303188323975, "num_tokens": 147271575.0, "step": 3859 }, { "epoch": 0.49103167535936904, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.963985443115234, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8547568321228027, "num_tokens": 147315473.0, "step": 3860 }, { "epoch": 0.49115888563795956, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.937638282775879, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.86150062084198, "num_tokens": 147359058.0, "step": 3861 }, { "epoch": 0.49128609591655004, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.971856117248535, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8633283972740173, "num_tokens": 147394443.0, "step": 3862 }, { "epoch": 0.49141330619514056, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.927600860595703, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8523213863372803, "num_tokens": 147436225.0, "step": 3863 }, { "epoch": 0.4915405164737311, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.97427749633789, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8677686452865601, "num_tokens": 147473449.0, "step": 3864 }, { "epoch": 0.49166772675232157, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.975317001342773, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8560509085655212, "num_tokens": 147510330.0, "step": 3865 }, { "epoch": 0.4917949370309121, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.93681812286377, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8698354363441467, "num_tokens": 147544683.0, "step": 3866 }, { "epoch": 0.4919221473095026, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.972578048706055, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8755578398704529, "num_tokens": 147589330.0, "step": 3867 }, { "epoch": 0.4920493575880931, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.924198150634766, "learning_rate": 1e-06, "loss": 0.5448, "mean_token_accuracy": 0.8271946907043457, "num_tokens": 147626864.0, "step": 3868 }, { "epoch": 0.4921765678666836, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.89163875579834, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.839821457862854, "num_tokens": 147672006.0, "step": 3869 }, { "epoch": 0.49230377814527415, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.94667911529541, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8484720587730408, "num_tokens": 147717388.0, "step": 3870 }, { "epoch": 0.4924309884238646, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.912214279174805, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8463084101676941, "num_tokens": 147759827.0, "step": 3871 }, { "epoch": 0.49255819870245515, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.93272590637207, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8391848802566528, "num_tokens": 147793742.0, "step": 3872 }, { "epoch": 0.4926854089810457, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.879212379455566, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8683226108551025, "num_tokens": 147829881.0, "step": 3873 }, { "epoch": 0.49281261925963615, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.922334671020508, "learning_rate": 1e-06, "loss": 0.4859, "mean_token_accuracy": 0.8432201743125916, "num_tokens": 147875051.0, "step": 3874 }, { "epoch": 0.4929398295382267, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.927772521972656, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.850859522819519, "num_tokens": 147912786.0, "step": 3875 }, { "epoch": 0.4930670398168172, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.959959030151367, "learning_rate": 1e-06, "loss": 0.5044, "mean_token_accuracy": 0.8367953300476074, "num_tokens": 147954981.0, "step": 3876 }, { "epoch": 0.4931942500954077, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.978655815124512, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8505610823631287, "num_tokens": 147992179.0, "step": 3877 }, { "epoch": 0.4933214603739982, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.919720649719238, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8629642724990845, "num_tokens": 148029866.0, "step": 3878 }, { "epoch": 0.49344867065258874, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.960358619689941, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8579939007759094, "num_tokens": 148070057.0, "step": 3879 }, { "epoch": 0.4935758809311792, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.982000350952148, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8612135648727417, "num_tokens": 148108461.0, "step": 3880 }, { "epoch": 0.49370309120976974, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.960539817810059, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8503520488739014, "num_tokens": 148147609.0, "step": 3881 }, { "epoch": 0.49383030148836027, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.972654342651367, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8518431782722473, "num_tokens": 148190475.0, "step": 3882 }, { "epoch": 0.4939575117669508, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.997177124023438, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8695563673973083, "num_tokens": 148221187.0, "step": 3883 }, { "epoch": 0.49408472204554127, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.994034767150879, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8660604357719421, "num_tokens": 148260916.0, "step": 3884 }, { "epoch": 0.4942119323241318, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.907856941223145, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8629473447799683, "num_tokens": 148295591.0, "step": 3885 }, { "epoch": 0.4943391426027223, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.954514503479004, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8526231050491333, "num_tokens": 148333445.0, "step": 3886 }, { "epoch": 0.4944663528813128, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.99662971496582, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.866097629070282, "num_tokens": 148367951.0, "step": 3887 }, { "epoch": 0.4945935631599033, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.979873657226562, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8496692180633545, "num_tokens": 148408491.0, "step": 3888 }, { "epoch": 0.49472077343849385, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.964361190795898, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8570159077644348, "num_tokens": 148443883.0, "step": 3889 }, { "epoch": 0.4948479837170843, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.977058410644531, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8589207530021667, "num_tokens": 148480511.0, "step": 3890 }, { "epoch": 0.49497519399567486, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.943428993225098, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8528196811676025, "num_tokens": 148514759.0, "step": 3891 }, { "epoch": 0.4951024042742654, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.895867347717285, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8594110012054443, "num_tokens": 148553209.0, "step": 3892 }, { "epoch": 0.49522961455285586, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.004557609558105, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8489137887954712, "num_tokens": 148590387.0, "step": 3893 }, { "epoch": 0.4953568248314464, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.987780570983887, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8553968071937561, "num_tokens": 148625201.0, "step": 3894 }, { "epoch": 0.4954840351100369, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.92296314239502, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8609609603881836, "num_tokens": 148665989.0, "step": 3895 }, { "epoch": 0.4956112453886274, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.011337280273438, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8606453537940979, "num_tokens": 148704839.0, "step": 3896 }, { "epoch": 0.4957384556672179, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.976957321166992, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8671562671661377, "num_tokens": 148740523.0, "step": 3897 }, { "epoch": 0.49586566594580844, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.963449478149414, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.873058557510376, "num_tokens": 148779103.0, "step": 3898 }, { "epoch": 0.4959928762243989, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.960783958435059, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8684632778167725, "num_tokens": 148814659.0, "step": 3899 }, { "epoch": 0.49612008650298944, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.936481475830078, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8780558109283447, "num_tokens": 148856864.0, "step": 3900 }, { "epoch": 0.49624729678157997, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.045671463012695, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8640027046203613, "num_tokens": 148890309.0, "step": 3901 }, { "epoch": 0.49637450706017044, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.02347183227539, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.8428983688354492, "num_tokens": 148926662.0, "step": 3902 }, { "epoch": 0.496501717338761, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.964313507080078, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8456661701202393, "num_tokens": 148963032.0, "step": 3903 }, { "epoch": 0.4966289276173515, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.040556907653809, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8633977174758911, "num_tokens": 148996684.0, "step": 3904 }, { "epoch": 0.496756137895942, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.956267356872559, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8583678603172302, "num_tokens": 149032735.0, "step": 3905 }, { "epoch": 0.4968833481745325, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.989442825317383, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8599793314933777, "num_tokens": 149069415.0, "step": 3906 }, { "epoch": 0.49701055845312303, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.000221252441406, "learning_rate": 1e-06, "loss": 0.491, "mean_token_accuracy": 0.8442103862762451, "num_tokens": 149115042.0, "step": 3907 }, { "epoch": 0.4971377687317135, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.054523468017578, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8662639260292053, "num_tokens": 149149767.0, "step": 3908 }, { "epoch": 0.49726497901030403, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.95926570892334, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8511513471603394, "num_tokens": 149193000.0, "step": 3909 }, { "epoch": 0.49739218928889456, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.032967567443848, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8546774387359619, "num_tokens": 149226963.0, "step": 3910 }, { "epoch": 0.49751939956748503, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.02283000946045, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.853087306022644, "num_tokens": 149268139.0, "step": 3911 }, { "epoch": 0.49764660984607556, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.974205017089844, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.864936351776123, "num_tokens": 149305882.0, "step": 3912 }, { "epoch": 0.4977738201246661, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.962936401367188, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8551053404808044, "num_tokens": 149344119.0, "step": 3913 }, { "epoch": 0.49790103040325656, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.026129722595215, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.848436176776886, "num_tokens": 149384102.0, "step": 3914 }, { "epoch": 0.4980282406818471, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.960052490234375, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8553352355957031, "num_tokens": 149425677.0, "step": 3915 }, { "epoch": 0.4981554509604376, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.991572380065918, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8546883463859558, "num_tokens": 149463275.0, "step": 3916 }, { "epoch": 0.4982826612390281, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.940428733825684, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8722318410873413, "num_tokens": 149503950.0, "step": 3917 }, { "epoch": 0.4984098715176186, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.985148429870605, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8489158153533936, "num_tokens": 149541192.0, "step": 3918 }, { "epoch": 0.49853708179620915, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.03089427947998, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8543111085891724, "num_tokens": 149581784.0, "step": 3919 }, { "epoch": 0.4986642920747996, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.99381160736084, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8631619215011597, "num_tokens": 149619312.0, "step": 3920 }, { "epoch": 0.49879150235339015, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.996837615966797, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8553459048271179, "num_tokens": 149657443.0, "step": 3921 }, { "epoch": 0.4989187126319807, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.010144233703613, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8687752485275269, "num_tokens": 149694992.0, "step": 3922 }, { "epoch": 0.49904592291057115, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.028751373291016, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8544372916221619, "num_tokens": 149729554.0, "step": 3923 }, { "epoch": 0.4991731331891617, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.994648933410645, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8634070158004761, "num_tokens": 149770051.0, "step": 3924 }, { "epoch": 0.4993003434677522, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.009822845458984, "learning_rate": 1e-06, "loss": 0.5387, "mean_token_accuracy": 0.826491117477417, "num_tokens": 149806724.0, "step": 3925 }, { "epoch": 0.4994275537463427, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.983498573303223, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.849689781665802, "num_tokens": 149849874.0, "step": 3926 }, { "epoch": 0.4995547640249332, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.992286682128906, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8571822643280029, "num_tokens": 149888076.0, "step": 3927 }, { "epoch": 0.49968197430352373, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.004334449768066, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.858837366104126, "num_tokens": 149923750.0, "step": 3928 }, { "epoch": 0.4998091845821142, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.012109756469727, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8452364802360535, "num_tokens": 149960657.0, "step": 3929 }, { "epoch": 0.49993639486070474, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.017923355102539, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8730156421661377, "num_tokens": 149995983.0, "step": 3930 }, { "epoch": 0.5000636051392953, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.957029342651367, "learning_rate": 1e-06, "loss": 0.5052, "mean_token_accuracy": 0.8358575701713562, "num_tokens": 150037508.0, "step": 3931 }, { "epoch": 0.5001908154178858, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.963573455810547, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8531954288482666, "num_tokens": 150077806.0, "step": 3932 }, { "epoch": 0.5003180256964763, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.991310119628906, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8557538390159607, "num_tokens": 150117589.0, "step": 3933 }, { "epoch": 0.5004452359750667, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.973779678344727, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8501933813095093, "num_tokens": 150161579.0, "step": 3934 }, { "epoch": 0.5005724462536573, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.998501777648926, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8448362350463867, "num_tokens": 150208018.0, "step": 3935 }, { "epoch": 0.5006996565322478, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.01400089263916, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8584970235824585, "num_tokens": 150250164.0, "step": 3936 }, { "epoch": 0.5008268668108383, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.026577949523926, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8760101199150085, "num_tokens": 150289972.0, "step": 3937 }, { "epoch": 0.5009540770894289, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.960168838500977, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8550962209701538, "num_tokens": 150333656.0, "step": 3938 }, { "epoch": 0.5010812873680194, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.01695728302002, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8705011010169983, "num_tokens": 150374086.0, "step": 3939 }, { "epoch": 0.5012084976466098, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.97969913482666, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8613690733909607, "num_tokens": 150407472.0, "step": 3940 }, { "epoch": 0.5013357079252003, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.023845672607422, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8633136749267578, "num_tokens": 150444560.0, "step": 3941 }, { "epoch": 0.5014629182037909, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.96257209777832, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.860813319683075, "num_tokens": 150483621.0, "step": 3942 }, { "epoch": 0.5015901284823814, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 10.042860984802246, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8574920296669006, "num_tokens": 150520414.0, "step": 3943 }, { "epoch": 0.5017173387609719, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.022934913635254, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8488898277282715, "num_tokens": 150556589.0, "step": 3944 }, { "epoch": 0.5018445490395624, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 10.025459289550781, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8557975888252258, "num_tokens": 150592393.0, "step": 3945 }, { "epoch": 0.5019717593181529, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.984615325927734, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.8442734479904175, "num_tokens": 150630457.0, "step": 3946 }, { "epoch": 0.5020989695967434, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 10.003358840942383, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8655228018760681, "num_tokens": 150659482.0, "step": 3947 }, { "epoch": 0.5022261798753339, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.971793174743652, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8575229644775391, "num_tokens": 150700920.0, "step": 3948 }, { "epoch": 0.5023533901539244, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.0531158447265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.983576774597168, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8689682483673096, "num_tokens": 150736393.0, "step": 3949 }, { "epoch": 0.502480600432515, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.940731048583984, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8547595739364624, "num_tokens": 150783848.0, "step": 3950 }, { "epoch": 0.5026078107111055, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.988303184509277, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8544173836708069, "num_tokens": 150824179.0, "step": 3951 }, { "epoch": 0.5027350209896959, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.025983810424805, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.857719361782074, "num_tokens": 150862775.0, "step": 3952 }, { "epoch": 0.5028622312682864, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.991040229797363, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8574919700622559, "num_tokens": 150900005.0, "step": 3953 }, { "epoch": 0.502989441546877, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 10.01485538482666, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8539823889732361, "num_tokens": 150936148.0, "step": 3954 }, { "epoch": 0.5031166518254675, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.0013427734375, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8542965650558472, "num_tokens": 150971661.0, "step": 3955 }, { "epoch": 0.503243862104058, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.985197067260742, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8687946200370789, "num_tokens": 151009521.0, "step": 3956 }, { "epoch": 0.5033710723826486, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.978038787841797, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.854217529296875, "num_tokens": 151053641.0, "step": 3957 }, { "epoch": 0.5034982826612391, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.045059204101562, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.861059844493866, "num_tokens": 151094476.0, "step": 3958 }, { "epoch": 0.5036254929398295, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.990193367004395, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8510582447052002, "num_tokens": 151126381.0, "step": 3959 }, { "epoch": 0.50375270321842, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.042662620544434, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.861312210559845, "num_tokens": 151156875.0, "step": 3960 }, { "epoch": 0.5038799134970106, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.929057121276855, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8528211712837219, "num_tokens": 151194394.0, "step": 3961 }, { "epoch": 0.5040071237756011, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.958732604980469, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8433732986450195, "num_tokens": 151234472.0, "step": 3962 }, { "epoch": 0.5041343340541916, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.070968627929688, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8530121445655823, "num_tokens": 151270426.0, "step": 3963 }, { "epoch": 0.5042615443327821, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.994611740112305, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8551768660545349, "num_tokens": 151305793.0, "step": 3964 }, { "epoch": 0.5043887546113726, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.00232982635498, "learning_rate": 1e-06, "loss": 0.4804, "mean_token_accuracy": 0.8447748422622681, "num_tokens": 151345116.0, "step": 3965 }, { "epoch": 0.5045159648899631, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.0531158447265625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.04818058013916, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8617662191390991, "num_tokens": 151381901.0, "step": 3966 }, { "epoch": 0.5046431751685536, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.9976224899292, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8587380051612854, "num_tokens": 151426884.0, "step": 3967 }, { "epoch": 0.5047703854471441, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.070150375366211, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8502559661865234, "num_tokens": 151467596.0, "step": 3968 }, { "epoch": 0.5048975957257347, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.022710800170898, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8766400218009949, "num_tokens": 151502131.0, "step": 3969 }, { "epoch": 0.5050248060043252, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.015280723571777, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8545504808425903, "num_tokens": 151538932.0, "step": 3970 }, { "epoch": 0.5051520162829156, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.981945991516113, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8561417460441589, "num_tokens": 151577839.0, "step": 3971 }, { "epoch": 0.5052792265615061, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.99923324584961, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8575731515884399, "num_tokens": 151618984.0, "step": 3972 }, { "epoch": 0.5054064368400967, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.02496337890625, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8572413921356201, "num_tokens": 151658392.0, "step": 3973 }, { "epoch": 0.5055336471186872, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.991219520568848, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8674747943878174, "num_tokens": 151689781.0, "step": 3974 }, { "epoch": 0.5056608573972777, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.059025764465332, "learning_rate": 1e-06, "loss": 0.5006, "mean_token_accuracy": 0.8425692319869995, "num_tokens": 151729612.0, "step": 3975 }, { "epoch": 0.5057880676758683, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.99932861328125, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8442862033843994, "num_tokens": 151769270.0, "step": 3976 }, { "epoch": 0.5059152779544587, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.05915641784668, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.8386016488075256, "num_tokens": 151803936.0, "step": 3977 }, { "epoch": 0.5060424882330492, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.04090404510498, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8453813791275024, "num_tokens": 151840429.0, "step": 3978 }, { "epoch": 0.5061696985116397, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.029082298278809, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8584307432174683, "num_tokens": 151883296.0, "step": 3979 }, { "epoch": 0.5062969087902303, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.019573211669922, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.872160792350769, "num_tokens": 151917658.0, "step": 3980 }, { "epoch": 0.5064241190688208, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.981059074401855, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8568062782287598, "num_tokens": 151958822.0, "step": 3981 }, { "epoch": 0.5065513293474113, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.011575698852539, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.851932942867279, "num_tokens": 151994964.0, "step": 3982 }, { "epoch": 0.5066785396260017, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.027371406555176, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8521654605865479, "num_tokens": 152037013.0, "step": 3983 }, { "epoch": 0.5068057499045923, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.99195384979248, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8527888059616089, "num_tokens": 152074821.0, "step": 3984 }, { "epoch": 0.5069329601831828, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.993239402770996, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8492825031280518, "num_tokens": 152116667.0, "step": 3985 }, { "epoch": 0.5070601704617733, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.001301765441895, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8607445955276489, "num_tokens": 152152714.0, "step": 3986 }, { "epoch": 0.5071873807403638, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.062479019165039, "learning_rate": 1e-06, "loss": 0.4899, "mean_token_accuracy": 0.8422666192054749, "num_tokens": 152190230.0, "step": 3987 }, { "epoch": 0.5073145910189544, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.996021270751953, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8560706377029419, "num_tokens": 152228887.0, "step": 3988 }, { "epoch": 0.5074418012975448, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.999164581298828, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.865587592124939, "num_tokens": 152271863.0, "step": 3989 }, { "epoch": 0.5075690115761353, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.011749267578125, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8524960875511169, "num_tokens": 152310765.0, "step": 3990 }, { "epoch": 0.5076962218547258, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.058387756347656, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8793023228645325, "num_tokens": 152342545.0, "step": 3991 }, { "epoch": 0.5078234321333164, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.012019157409668, "learning_rate": 1e-06, "loss": 0.4872, "mean_token_accuracy": 0.8409588932991028, "num_tokens": 152386223.0, "step": 3992 }, { "epoch": 0.5079506424119069, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.026611328125, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8635099530220032, "num_tokens": 152421811.0, "step": 3993 }, { "epoch": 0.5080778526904974, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.061725616455078, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8447950482368469, "num_tokens": 152454581.0, "step": 3994 }, { "epoch": 0.5082050629690879, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.041762351989746, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8469270467758179, "num_tokens": 152490591.0, "step": 3995 }, { "epoch": 0.5083322732476784, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.038167953491211, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8799065351486206, "num_tokens": 152532290.0, "step": 3996 }, { "epoch": 0.5084594835262689, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.070921897888184, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8619639873504639, "num_tokens": 152571984.0, "step": 3997 }, { "epoch": 0.5085866938048594, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.092966079711914, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8569895029067993, "num_tokens": 152611082.0, "step": 3998 }, { "epoch": 0.50871390408345, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.963539123535156, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8646628856658936, "num_tokens": 152655303.0, "step": 3999 }, { "epoch": 0.5088411143620405, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.135845184326172, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8557164072990417, "num_tokens": 152688432.0, "step": 4000 }, { "epoch": 0.5089683246406309, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.065987586975098, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8532495498657227, "num_tokens": 152725637.0, "step": 4001 }, { "epoch": 0.5090955349192214, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.993457794189453, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8548606634140015, "num_tokens": 152763108.0, "step": 4002 }, { "epoch": 0.509222745197812, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.048126220703125, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8635520339012146, "num_tokens": 152800659.0, "step": 4003 }, { "epoch": 0.5093499554764025, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.088675498962402, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.848074197769165, "num_tokens": 152842797.0, "step": 4004 }, { "epoch": 0.509477165754993, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.01874828338623, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.84744793176651, "num_tokens": 152888257.0, "step": 4005 }, { "epoch": 0.5096043760335836, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.044450759887695, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8562451004981995, "num_tokens": 152929089.0, "step": 4006 }, { "epoch": 0.5097315863121741, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.047149658203125, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8502859473228455, "num_tokens": 152972710.0, "step": 4007 }, { "epoch": 0.5098587965907645, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.096505165100098, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8570646047592163, "num_tokens": 153008867.0, "step": 4008 }, { "epoch": 0.509986006869355, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.090962409973145, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8518747687339783, "num_tokens": 153051497.0, "step": 4009 }, { "epoch": 0.5101132171479456, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.013219833374023, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8733339309692383, "num_tokens": 153089295.0, "step": 4010 }, { "epoch": 0.5102404274265361, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.098438262939453, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8616799116134644, "num_tokens": 153123957.0, "step": 4011 }, { "epoch": 0.5103676377051266, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.070043563842773, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8536967039108276, "num_tokens": 153165248.0, "step": 4012 }, { "epoch": 0.5104948479837171, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.027923583984375, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8715275526046753, "num_tokens": 153206163.0, "step": 4013 }, { "epoch": 0.5106220582623076, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.040776252746582, "learning_rate": 1e-06, "loss": 0.4728, "mean_token_accuracy": 0.8477142453193665, "num_tokens": 153248666.0, "step": 4014 }, { "epoch": 0.5107492685408981, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.000434875488281, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8797175884246826, "num_tokens": 153289009.0, "step": 4015 }, { "epoch": 0.5108764788194886, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.052268028259277, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.862628161907196, "num_tokens": 153321354.0, "step": 4016 }, { "epoch": 0.5110036890980791, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.03846549987793, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8564648628234863, "num_tokens": 153361061.0, "step": 4017 }, { "epoch": 0.5111308993766697, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.100799560546875e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.043193817138672, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8509201407432556, "num_tokens": 153397002.0, "step": 4018 }, { "epoch": 0.5112581096552602, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.067941665649414, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8439752459526062, "num_tokens": 153435264.0, "step": 4019 }, { "epoch": 0.5113853199338506, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.047805786132812, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8551781177520752, "num_tokens": 153474751.0, "step": 4020 }, { "epoch": 0.5115125302124411, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.993483543395996, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8667306900024414, "num_tokens": 153510084.0, "step": 4021 }, { "epoch": 0.5116397404910317, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 10.001351356506348, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8436862230300903, "num_tokens": 153550387.0, "step": 4022 }, { "epoch": 0.5117669507696222, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.02774429321289, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8789975047111511, "num_tokens": 153592441.0, "step": 4023 }, { "epoch": 0.5118941610482127, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.981101989746094, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8642358779907227, "num_tokens": 153629202.0, "step": 4024 }, { "epoch": 0.5120213713268033, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.028220176696777, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8525602221488953, "num_tokens": 153669243.0, "step": 4025 }, { "epoch": 0.5121485816053937, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.991185188293457, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8525017499923706, "num_tokens": 153709206.0, "step": 4026 }, { "epoch": 0.5122757918839842, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.068581581115723, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8633650541305542, "num_tokens": 153741251.0, "step": 4027 }, { "epoch": 0.5124030021625747, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.997468948364258, "learning_rate": 1e-06, "loss": 0.4929, "mean_token_accuracy": 0.8431926369667053, "num_tokens": 153775559.0, "step": 4028 }, { "epoch": 0.5125302124411653, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.034207344055176, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8611456751823425, "num_tokens": 153819158.0, "step": 4029 }, { "epoch": 0.5126574227197558, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.97867488861084, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8514681458473206, "num_tokens": 153856231.0, "step": 4030 }, { "epoch": 0.5127846329983463, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.020535469055176, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8681931495666504, "num_tokens": 153897125.0, "step": 4031 }, { "epoch": 0.5129118432769367, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.988056182861328, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8645192384719849, "num_tokens": 153933552.0, "step": 4032 }, { "epoch": 0.5130390535555273, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.04846477508545, "learning_rate": 1e-06, "loss": 0.4967, "mean_token_accuracy": 0.8425424695014954, "num_tokens": 153974337.0, "step": 4033 }, { "epoch": 0.5131662638341178, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.001710891723633, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8521144986152649, "num_tokens": 154018911.0, "step": 4034 }, { "epoch": 0.5132934741127083, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.053559303283691, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.867390513420105, "num_tokens": 154056441.0, "step": 4035 }, { "epoch": 0.5134206843912988, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.090085983276367, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8475380539894104, "num_tokens": 154096326.0, "step": 4036 }, { "epoch": 0.5135478946698894, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.082854270935059, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8608129620552063, "num_tokens": 154137223.0, "step": 4037 }, { "epoch": 0.5136751049484798, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.107061386108398, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8552930951118469, "num_tokens": 154168312.0, "step": 4038 }, { "epoch": 0.5138023152270703, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.09683895111084, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8709992170333862, "num_tokens": 154203125.0, "step": 4039 }, { "epoch": 0.5139295255056608, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.041596412658691, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8512620329856873, "num_tokens": 154248437.0, "step": 4040 }, { "epoch": 0.5140567357842514, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.112607955932617, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8614624738693237, "num_tokens": 154286303.0, "step": 4041 }, { "epoch": 0.5141839460628419, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.081171989440918, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8480697274208069, "num_tokens": 154328638.0, "step": 4042 }, { "epoch": 0.5143111563414324, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.072040557861328, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8557267785072327, "num_tokens": 154362953.0, "step": 4043 }, { "epoch": 0.5144383666200228, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.100358009338379, "learning_rate": 1e-06, "loss": 0.4897, "mean_token_accuracy": 0.8449912071228027, "num_tokens": 154401289.0, "step": 4044 }, { "epoch": 0.5145655768986134, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.121651649475098, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8608443140983582, "num_tokens": 154438779.0, "step": 4045 }, { "epoch": 0.5146927871772039, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.110577583312988, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8511788845062256, "num_tokens": 154467032.0, "step": 4046 }, { "epoch": 0.5148199974557944, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.063213348388672, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8668427467346191, "num_tokens": 154505222.0, "step": 4047 }, { "epoch": 0.514947207734385, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.059640884399414, "learning_rate": 1e-06, "loss": 0.4999, "mean_token_accuracy": 0.8438467979431152, "num_tokens": 154544953.0, "step": 4048 }, { "epoch": 0.5150744180129755, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.021279335021973, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8692100048065186, "num_tokens": 154586193.0, "step": 4049 }, { "epoch": 0.5152016282915659, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.023541450500488, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8611959218978882, "num_tokens": 154629531.0, "step": 4050 }, { "epoch": 0.5153288385701564, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.061586380004883, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8724949955940247, "num_tokens": 154659482.0, "step": 4051 }, { "epoch": 0.515456048848747, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.065878868103027, "learning_rate": 1e-06, "loss": 0.5384, "mean_token_accuracy": 0.8242601156234741, "num_tokens": 154702645.0, "step": 4052 }, { "epoch": 0.5155832591273375, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.00958251953125, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8653194308280945, "num_tokens": 154743329.0, "step": 4053 }, { "epoch": 0.515710469405928, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.019508361816406, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.853971004486084, "num_tokens": 154783460.0, "step": 4054 }, { "epoch": 0.5158376796845185, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.05553150177002, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8507996797561646, "num_tokens": 154827209.0, "step": 4055 }, { "epoch": 0.5159648899631091, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.08513069152832, "learning_rate": 1e-06, "loss": 0.5202, "mean_token_accuracy": 0.8327199220657349, "num_tokens": 154866577.0, "step": 4056 }, { "epoch": 0.5160921002416995, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.009176254272461, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8593916296958923, "num_tokens": 154912191.0, "step": 4057 }, { "epoch": 0.51621931052029, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.064545631408691, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8436312675476074, "num_tokens": 154953412.0, "step": 4058 }, { "epoch": 0.5163465207988805, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.03661823272705, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8682960867881775, "num_tokens": 154989913.0, "step": 4059 }, { "epoch": 0.5164737310774711, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.050461769104004, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8441359996795654, "num_tokens": 155038094.0, "step": 4060 }, { "epoch": 0.5166009413560616, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.070014953613281, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.847201943397522, "num_tokens": 155075900.0, "step": 4061 }, { "epoch": 0.5167281516346521, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.047067642211914, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8557668924331665, "num_tokens": 155117038.0, "step": 4062 }, { "epoch": 0.5168553619132426, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.045695304870605, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8718302845954895, "num_tokens": 155151442.0, "step": 4063 }, { "epoch": 0.5169825721918331, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.086165428161621, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.865079402923584, "num_tokens": 155192288.0, "step": 4064 }, { "epoch": 0.5171097824704236, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.030618667602539, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.858906090259552, "num_tokens": 155238931.0, "step": 4065 }, { "epoch": 0.5172369927490141, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.018996238708496, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8701839447021484, "num_tokens": 155276644.0, "step": 4066 }, { "epoch": 0.5173642030276047, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.108159065246582, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8692439794540405, "num_tokens": 155307638.0, "step": 4067 }, { "epoch": 0.5174914133061952, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.1111421585083, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8508954048156738, "num_tokens": 155343255.0, "step": 4068 }, { "epoch": 0.5176186235847856, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.107083320617676, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8588727712631226, "num_tokens": 155382283.0, "step": 4069 }, { "epoch": 0.5177458338633761, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.085901260375977, "learning_rate": 1e-06, "loss": 0.5121, "mean_token_accuracy": 0.8352490067481995, "num_tokens": 155427096.0, "step": 4070 }, { "epoch": 0.5178730441419667, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.059942245483398, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8452714085578918, "num_tokens": 155467472.0, "step": 4071 }, { "epoch": 0.5180002544205572, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.060127258300781, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8516610860824585, "num_tokens": 155508350.0, "step": 4072 }, { "epoch": 0.5181274646991477, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.080740928649902, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8806874752044678, "num_tokens": 155547215.0, "step": 4073 }, { "epoch": 0.5182546749777382, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.057056427001953, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.861964225769043, "num_tokens": 155586658.0, "step": 4074 }, { "epoch": 0.5183818852563287, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.092373847961426, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8594197630882263, "num_tokens": 155620944.0, "step": 4075 }, { "epoch": 0.5185090955349192, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.113187789916992, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8540290594100952, "num_tokens": 155662699.0, "step": 4076 }, { "epoch": 0.5186363058135097, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.07335090637207, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8450766205787659, "num_tokens": 155707783.0, "step": 4077 }, { "epoch": 0.5187635160921003, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.051492691040039, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8628199100494385, "num_tokens": 155745227.0, "step": 4078 }, { "epoch": 0.5188907263706908, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.075974464416504, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8585178852081299, "num_tokens": 155786466.0, "step": 4079 }, { "epoch": 0.5190179366492813, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.030488967895508, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8588305711746216, "num_tokens": 155833145.0, "step": 4080 }, { "epoch": 0.5191451469278717, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.085408210754395, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8474289178848267, "num_tokens": 155869533.0, "step": 4081 }, { "epoch": 0.5192723572064623, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.111506462097168, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.856013298034668, "num_tokens": 155909329.0, "step": 4082 }, { "epoch": 0.5193995674850528, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.10062313079834, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8656594753265381, "num_tokens": 155947848.0, "step": 4083 }, { "epoch": 0.5195267777636433, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.061761856079102, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.861238420009613, "num_tokens": 155982693.0, "step": 4084 }, { "epoch": 0.5196539880422338, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.107795715332031, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8565056324005127, "num_tokens": 156023414.0, "step": 4085 }, { "epoch": 0.5197811983208244, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.089866638183594, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8566579818725586, "num_tokens": 156063488.0, "step": 4086 }, { "epoch": 0.5199084085994148, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.040412902832031, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8571361899375916, "num_tokens": 156103081.0, "step": 4087 }, { "epoch": 0.5200356188780053, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.074950218200684, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8491381406784058, "num_tokens": 156139453.0, "step": 4088 }, { "epoch": 0.5201628291565958, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.048964500427246, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8520079255104065, "num_tokens": 156176564.0, "step": 4089 }, { "epoch": 0.5202900394351864, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.052044868469238, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8600976467132568, "num_tokens": 156221278.0, "step": 4090 }, { "epoch": 0.5204172497137769, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.031676292419434, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8620114326477051, "num_tokens": 156259124.0, "step": 4091 }, { "epoch": 0.5205444599923674, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.0055570602417, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.847070574760437, "num_tokens": 156297832.0, "step": 4092 }, { "epoch": 0.5206716702709578, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.071626663208008, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.862926721572876, "num_tokens": 156329745.0, "step": 4093 }, { "epoch": 0.5207988805495484, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.06246566772461, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8575751185417175, "num_tokens": 156360154.0, "step": 4094 }, { "epoch": 0.5209260908281389, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.039414405822754, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8605201244354248, "num_tokens": 156399360.0, "step": 4095 }, { "epoch": 0.5210533011067294, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.016963958740234, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.856379508972168, "num_tokens": 156438782.0, "step": 4096 }, { "epoch": 0.52118051138532, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.120532989501953, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.863537073135376, "num_tokens": 156470261.0, "step": 4097 }, { "epoch": 0.5213077216639105, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.01642894744873, "learning_rate": 1e-06, "loss": 0.4984, "mean_token_accuracy": 0.8403713703155518, "num_tokens": 156510493.0, "step": 4098 }, { "epoch": 0.5214349319425009, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.057173728942871, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8595893383026123, "num_tokens": 156545040.0, "step": 4099 }, { "epoch": 0.5215621422210914, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.08657169342041, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.855304479598999, "num_tokens": 156583474.0, "step": 4100 }, { "epoch": 0.521689352499682, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.025096893310547, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8567751049995422, "num_tokens": 156627137.0, "step": 4101 }, { "epoch": 0.5218165627782725, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.064408302307129, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.8460366725921631, "num_tokens": 156664726.0, "step": 4102 }, { "epoch": 0.521943773056863, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.037391662597656, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8750345706939697, "num_tokens": 156707014.0, "step": 4103 }, { "epoch": 0.5220709833354535, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.013224601745605, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8539285659790039, "num_tokens": 156747784.0, "step": 4104 }, { "epoch": 0.522198193614044, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.083961486816406, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8553178906440735, "num_tokens": 156784760.0, "step": 4105 }, { "epoch": 0.5223254038926345, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.091313362121582, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8543745279312134, "num_tokens": 156822899.0, "step": 4106 }, { "epoch": 0.522452614171225, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.056715965270996, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8638750910758972, "num_tokens": 156861308.0, "step": 4107 }, { "epoch": 0.5225798244498155, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.090679168701172, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8643448352813721, "num_tokens": 156901981.0, "step": 4108 }, { "epoch": 0.5227070347284061, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.099381446838379, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8630567789077759, "num_tokens": 156942215.0, "step": 4109 }, { "epoch": 0.5228342450069966, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.046464920043945, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8643213510513306, "num_tokens": 156983820.0, "step": 4110 }, { "epoch": 0.5229614552855871, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.118532180786133, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8699445128440857, "num_tokens": 157016215.0, "step": 4111 }, { "epoch": 0.5230886655641775, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.15213680267334, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8570919036865234, "num_tokens": 157052071.0, "step": 4112 }, { "epoch": 0.5232158758427681, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.068263053894043, "learning_rate": 1e-06, "loss": 0.5151, "mean_token_accuracy": 0.8359690308570862, "num_tokens": 157097711.0, "step": 4113 }, { "epoch": 0.5233430861213586, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.095035552978516, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8558897972106934, "num_tokens": 157134801.0, "step": 4114 }, { "epoch": 0.5234702963999491, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.134126663208008, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8573620319366455, "num_tokens": 157171885.0, "step": 4115 }, { "epoch": 0.5235975066785397, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.094230651855469, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.855968713760376, "num_tokens": 157211915.0, "step": 4116 }, { "epoch": 0.5237247169571302, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.047234535217285, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8513327836990356, "num_tokens": 157260564.0, "step": 4117 }, { "epoch": 0.5238519272357206, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.104728698730469, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8591058254241943, "num_tokens": 157294676.0, "step": 4118 }, { "epoch": 0.5239791375143111, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.04899787902832, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8677407503128052, "num_tokens": 157334607.0, "step": 4119 }, { "epoch": 0.5241063477929017, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.035719871520996, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8593893647193909, "num_tokens": 157372021.0, "step": 4120 }, { "epoch": 0.5242335580714922, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.061966896057129, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8621100783348083, "num_tokens": 157410092.0, "step": 4121 }, { "epoch": 0.5243607683500827, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.07658576965332, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8547062873840332, "num_tokens": 157447979.0, "step": 4122 }, { "epoch": 0.5244879786286732, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.058934211730957, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8531526923179626, "num_tokens": 157481376.0, "step": 4123 }, { "epoch": 0.5246151889072637, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.104086875915527, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8633136749267578, "num_tokens": 157519555.0, "step": 4124 }, { "epoch": 0.5247423991858542, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.127472877502441, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8559035062789917, "num_tokens": 157551849.0, "step": 4125 }, { "epoch": 0.5248696094644447, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.056897163391113, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8575805425643921, "num_tokens": 157590823.0, "step": 4126 }, { "epoch": 0.5249968197430352, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.023585319519043, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8677690029144287, "num_tokens": 157633771.0, "step": 4127 }, { "epoch": 0.5251240300216258, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.107841491699219, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8708305358886719, "num_tokens": 157667796.0, "step": 4128 }, { "epoch": 0.5252512403002163, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.044289588928223, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8673582673072815, "num_tokens": 157709452.0, "step": 4129 }, { "epoch": 0.5253784505788067, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.136494636535645, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.849180281162262, "num_tokens": 157741407.0, "step": 4130 }, { "epoch": 0.5255056608573972, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.109485626220703, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8685475587844849, "num_tokens": 157777693.0, "step": 4131 }, { "epoch": 0.5256328711359878, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.046727180480957, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8563082218170166, "num_tokens": 157814729.0, "step": 4132 }, { "epoch": 0.5257600814145783, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.008846282958984, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8672926425933838, "num_tokens": 157853698.0, "step": 4133 }, { "epoch": 0.5258872916931688, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.097540855407715, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8629207611083984, "num_tokens": 157888580.0, "step": 4134 }, { "epoch": 0.5260145019717594, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.042572021484375, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8555827140808105, "num_tokens": 157928434.0, "step": 4135 }, { "epoch": 0.5261417122503498, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.101226806640625, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8533960580825806, "num_tokens": 157968936.0, "step": 4136 }, { "epoch": 0.5262689225289403, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.107522010803223, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.84715336561203, "num_tokens": 158005496.0, "step": 4137 }, { "epoch": 0.5263961328075308, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.102702140808105, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8655192852020264, "num_tokens": 158047569.0, "step": 4138 }, { "epoch": 0.5265233430861214, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.029900550842285, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8565456867218018, "num_tokens": 158085733.0, "step": 4139 }, { "epoch": 0.5266505533647119, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.059991836547852, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8479816913604736, "num_tokens": 158123906.0, "step": 4140 }, { "epoch": 0.5267777636433024, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.070632934570312, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8640124201774597, "num_tokens": 158163663.0, "step": 4141 }, { "epoch": 0.5269049739218928, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.0352783203125, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8461219072341919, "num_tokens": 158205878.0, "step": 4142 }, { "epoch": 0.5270321842004834, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.067117691040039, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8609040379524231, "num_tokens": 158241608.0, "step": 4143 }, { "epoch": 0.5271593944790739, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.147462844848633, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8705442547798157, "num_tokens": 158274259.0, "step": 4144 }, { "epoch": 0.5272866047576644, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.102397918701172, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8641024827957153, "num_tokens": 158304793.0, "step": 4145 }, { "epoch": 0.527413815036255, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.040900230407715, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8713752031326294, "num_tokens": 158349420.0, "step": 4146 }, { "epoch": 0.5275410253148455, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.092035293579102, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8569918274879456, "num_tokens": 158395834.0, "step": 4147 }, { "epoch": 0.5276682355934359, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.097040176391602, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8594551086425781, "num_tokens": 158435568.0, "step": 4148 }, { "epoch": 0.5277954458720264, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.134539604187012, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8683047294616699, "num_tokens": 158473584.0, "step": 4149 }, { "epoch": 0.527922656150617, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.061084747314453, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8588966131210327, "num_tokens": 158513435.0, "step": 4150 }, { "epoch": 0.5280498664292075, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.107771873474121, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8586474657058716, "num_tokens": 158549478.0, "step": 4151 }, { "epoch": 0.528177076707798, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.123444557189941, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8628526926040649, "num_tokens": 158598742.0, "step": 4152 }, { "epoch": 0.5283042869863885, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.05873966217041, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8463493585586548, "num_tokens": 158636806.0, "step": 4153 }, { "epoch": 0.528431497264979, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.114063262939453, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8516209125518799, "num_tokens": 158676043.0, "step": 4154 }, { "epoch": 0.5285587075435695, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.066932678222656, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8419758081436157, "num_tokens": 158711772.0, "step": 4155 }, { "epoch": 0.52868591782216, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.142705917358398, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8572103977203369, "num_tokens": 158749608.0, "step": 4156 }, { "epoch": 0.5288131281007505, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.078996658325195, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8591958284378052, "num_tokens": 158790502.0, "step": 4157 }, { "epoch": 0.5289403383793411, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.065390586853027, "learning_rate": 1e-06, "loss": 0.5149, "mean_token_accuracy": 0.8348246216773987, "num_tokens": 158828415.0, "step": 4158 }, { "epoch": 0.5290675486579316, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.140425682067871, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8488821983337402, "num_tokens": 158866972.0, "step": 4159 }, { "epoch": 0.5291947589365221, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.043251037597656, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8559930324554443, "num_tokens": 158902991.0, "step": 4160 }, { "epoch": 0.5293219692151125, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.0853271484375, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.865494430065155, "num_tokens": 158945963.0, "step": 4161 }, { "epoch": 0.5294491794937031, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.097078323364258, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8560823202133179, "num_tokens": 158985407.0, "step": 4162 }, { "epoch": 0.5295763897722936, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.11290454864502, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8510174751281738, "num_tokens": 159024387.0, "step": 4163 }, { "epoch": 0.5297036000508841, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.146766662597656, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8495010137557983, "num_tokens": 159061430.0, "step": 4164 }, { "epoch": 0.5298308103294747, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.098305702209473, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.859740138053894, "num_tokens": 159099350.0, "step": 4165 }, { "epoch": 0.5299580206080652, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.153828620910645, "learning_rate": 1e-06, "loss": 0.5304, "mean_token_accuracy": 0.832109808921814, "num_tokens": 159136276.0, "step": 4166 }, { "epoch": 0.5300852308866556, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.170995712280273, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8797492384910583, "num_tokens": 159177102.0, "step": 4167 }, { "epoch": 0.5302124411652461, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.092966079711914, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8690637946128845, "num_tokens": 159213582.0, "step": 4168 }, { "epoch": 0.5303396514438367, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.12398910522461, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8565714359283447, "num_tokens": 159254283.0, "step": 4169 }, { "epoch": 0.5304668617224272, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.149263381958008, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8592905402183533, "num_tokens": 159288557.0, "step": 4170 }, { "epoch": 0.5305940720010177, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.093184471130371, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8675795793533325, "num_tokens": 159325047.0, "step": 4171 }, { "epoch": 0.5307212822796082, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.177067756652832, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.861672580242157, "num_tokens": 159366291.0, "step": 4172 }, { "epoch": 0.5308484925581987, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.123696327209473, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8516980409622192, "num_tokens": 159401634.0, "step": 4173 }, { "epoch": 0.5309757028367892, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.125638961791992, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.849335789680481, "num_tokens": 159435316.0, "step": 4174 }, { "epoch": 0.5311029131153797, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.12191390991211, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8511489629745483, "num_tokens": 159472192.0, "step": 4175 }, { "epoch": 0.5312301233939702, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.052865028381348, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8648892641067505, "num_tokens": 159515342.0, "step": 4176 }, { "epoch": 0.5313573336725608, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.065488815307617, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8622751235961914, "num_tokens": 159553932.0, "step": 4177 }, { "epoch": 0.5314845439511513, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.133315086364746, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8594571352005005, "num_tokens": 159584711.0, "step": 4178 }, { "epoch": 0.5316117542297417, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.125732421875, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.844862163066864, "num_tokens": 159618784.0, "step": 4179 }, { "epoch": 0.5317389645083322, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.039076805114746, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8563241958618164, "num_tokens": 159661288.0, "step": 4180 }, { "epoch": 0.5318661747869228, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.090733528137207, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.869035542011261, "num_tokens": 159701286.0, "step": 4181 }, { "epoch": 0.5319933850655133, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.10998821258545, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8462522625923157, "num_tokens": 159746631.0, "step": 4182 }, { "epoch": 0.5321205953441038, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.062370300292969, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8617539405822754, "num_tokens": 159787753.0, "step": 4183 }, { "epoch": 0.5322478056226944, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.107905387878418, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8503607511520386, "num_tokens": 159823092.0, "step": 4184 }, { "epoch": 0.5323750159012848, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.099626541137695, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8496958613395691, "num_tokens": 159858480.0, "step": 4185 }, { "epoch": 0.5325022261798753, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.093268394470215, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8483282327651978, "num_tokens": 159896220.0, "step": 4186 }, { "epoch": 0.5326294364584658, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.077963829040527, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8555878400802612, "num_tokens": 159930995.0, "step": 4187 }, { "epoch": 0.5327566467370564, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.112527847290039, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8530731201171875, "num_tokens": 159968013.0, "step": 4188 }, { "epoch": 0.5328838570156469, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.074161529541016, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8503144383430481, "num_tokens": 160004085.0, "step": 4189 }, { "epoch": 0.5330110672942374, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.131988525390625, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8688035011291504, "num_tokens": 160040606.0, "step": 4190 }, { "epoch": 0.5331382775728278, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.068107604980469, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8695427775382996, "num_tokens": 160080792.0, "step": 4191 }, { "epoch": 0.5332654878514184, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.125533103942871, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8461991548538208, "num_tokens": 160117055.0, "step": 4192 }, { "epoch": 0.5333926981300089, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.135550498962402, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8629506826400757, "num_tokens": 160153223.0, "step": 4193 }, { "epoch": 0.5335199084085994, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.13521671295166, "learning_rate": 1e-06, "loss": 0.4978, "mean_token_accuracy": 0.8409539461135864, "num_tokens": 160190920.0, "step": 4194 }, { "epoch": 0.53364711868719, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.110359191894531, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8558470010757446, "num_tokens": 160221898.0, "step": 4195 }, { "epoch": 0.5337743289657805, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.114717483520508, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8601120710372925, "num_tokens": 160267102.0, "step": 4196 }, { "epoch": 0.5339015392443709, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.051386833190918, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8631007671356201, "num_tokens": 160304422.0, "step": 4197 }, { "epoch": 0.5340287495229614, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.12989616394043, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8543807864189148, "num_tokens": 160339006.0, "step": 4198 }, { "epoch": 0.534155959801552, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.053732872009277, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8770898580551147, "num_tokens": 160377504.0, "step": 4199 }, { "epoch": 0.5342831700801425, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.128657341003418, "learning_rate": 1e-06, "loss": 0.4979, "mean_token_accuracy": 0.8374887704849243, "num_tokens": 160413315.0, "step": 4200 }, { "epoch": 0.534410380358733, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.051552772521973, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8720948696136475, "num_tokens": 160455034.0, "step": 4201 }, { "epoch": 0.5345375906373235, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.144453048706055, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8527741432189941, "num_tokens": 160490505.0, "step": 4202 }, { "epoch": 0.534664800915914, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.046978950500488, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8685119152069092, "num_tokens": 160528712.0, "step": 4203 }, { "epoch": 0.5347920111945045, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.156060218811035, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8494071960449219, "num_tokens": 160567945.0, "step": 4204 }, { "epoch": 0.534919221473095, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.095977783203125, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8641283512115479, "num_tokens": 160596767.0, "step": 4205 }, { "epoch": 0.5350464317516855, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.057442665100098, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.854529619216919, "num_tokens": 160636912.0, "step": 4206 }, { "epoch": 0.5351736420302761, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.101879119873047, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8508204817771912, "num_tokens": 160674588.0, "step": 4207 }, { "epoch": 0.5353008523088666, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.105128288269043, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8460649847984314, "num_tokens": 160713441.0, "step": 4208 }, { "epoch": 0.5354280625874571, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.093927383422852, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.867857813835144, "num_tokens": 160752716.0, "step": 4209 }, { "epoch": 0.5355552728660475, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.045549392700195, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8682337999343872, "num_tokens": 160798771.0, "step": 4210 }, { "epoch": 0.5356824831446381, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.102920532226562, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8754116296768188, "num_tokens": 160843085.0, "step": 4211 }, { "epoch": 0.5358096934232286, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.067492485046387, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8510081171989441, "num_tokens": 160883073.0, "step": 4212 }, { "epoch": 0.5359369037018191, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.141016006469727, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8646646738052368, "num_tokens": 160923734.0, "step": 4213 }, { "epoch": 0.5360641139804097, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.090352058410645, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8542920351028442, "num_tokens": 160963460.0, "step": 4214 }, { "epoch": 0.5361913242590002, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.165555953979492, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8553914427757263, "num_tokens": 161003691.0, "step": 4215 }, { "epoch": 0.5363185345375906, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.055181503295898, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8582466840744019, "num_tokens": 161042056.0, "step": 4216 }, { "epoch": 0.5364457448161811, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.12026596069336, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8813096880912781, "num_tokens": 161078670.0, "step": 4217 }, { "epoch": 0.5365729550947717, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.120285987854004, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8711098432540894, "num_tokens": 161113992.0, "step": 4218 }, { "epoch": 0.5367001653733622, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.100774765014648, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8638905882835388, "num_tokens": 161153763.0, "step": 4219 }, { "epoch": 0.5368273756519527, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.178348541259766, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8432422876358032, "num_tokens": 161184186.0, "step": 4220 }, { "epoch": 0.5369545859305432, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.066081047058105, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8550639152526855, "num_tokens": 161219298.0, "step": 4221 }, { "epoch": 0.5370817962091337, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.106683731079102, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8737931251525879, "num_tokens": 161255252.0, "step": 4222 }, { "epoch": 0.5372090064877242, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.124114036560059, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.861850917339325, "num_tokens": 161295890.0, "step": 4223 }, { "epoch": 0.5373362167663147, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.161296844482422, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8559464812278748, "num_tokens": 161330808.0, "step": 4224 }, { "epoch": 0.5374634270449052, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.067465782165527, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8659878373146057, "num_tokens": 161370515.0, "step": 4225 }, { "epoch": 0.5375906373234958, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.165526390075684, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8704847097396851, "num_tokens": 161405257.0, "step": 4226 }, { "epoch": 0.5377178476020863, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.120805740356445, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8589218854904175, "num_tokens": 161443526.0, "step": 4227 }, { "epoch": 0.5378450578806767, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.184507369995117, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8521105051040649, "num_tokens": 161481262.0, "step": 4228 }, { "epoch": 0.5379722681592672, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.131099700927734, "learning_rate": 1e-06, "loss": 0.4832, "mean_token_accuracy": 0.8438796997070312, "num_tokens": 161517756.0, "step": 4229 }, { "epoch": 0.5380994784378578, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.148466110229492, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8672699332237244, "num_tokens": 161552680.0, "step": 4230 }, { "epoch": 0.5382266887164483, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.12238597869873, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8554309010505676, "num_tokens": 161588607.0, "step": 4231 }, { "epoch": 0.5383538989950388, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.140338897705078, "learning_rate": 1e-06, "loss": 0.4935, "mean_token_accuracy": 0.8425576090812683, "num_tokens": 161626427.0, "step": 4232 }, { "epoch": 0.5384811092736294, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.038747787475586, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8642389178276062, "num_tokens": 161665182.0, "step": 4233 }, { "epoch": 0.5386083195522198, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.14680004119873, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.859108567237854, "num_tokens": 161704685.0, "step": 4234 }, { "epoch": 0.5387355298308103, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.084310531616211, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8647723197937012, "num_tokens": 161742960.0, "step": 4235 }, { "epoch": 0.5388627401094008, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.103228569030762, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8577698469161987, "num_tokens": 161778199.0, "step": 4236 }, { "epoch": 0.5389899503879914, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.073854446411133, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8546342253684998, "num_tokens": 161811235.0, "step": 4237 }, { "epoch": 0.5391171606665819, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.161772727966309, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8588956594467163, "num_tokens": 161847669.0, "step": 4238 }, { "epoch": 0.5392443709451724, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.116876602172852, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.844953715801239, "num_tokens": 161888470.0, "step": 4239 }, { "epoch": 0.5393715812237628, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.132795333862305, "learning_rate": 1e-06, "loss": 0.4978, "mean_token_accuracy": 0.8397881388664246, "num_tokens": 161924563.0, "step": 4240 }, { "epoch": 0.5394987915023534, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.146334648132324, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8448485732078552, "num_tokens": 161959062.0, "step": 4241 }, { "epoch": 0.5396260017809439, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.109265327453613, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8628778457641602, "num_tokens": 161994755.0, "step": 4242 }, { "epoch": 0.5397532120595344, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.122060775756836, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8682788610458374, "num_tokens": 162032812.0, "step": 4243 }, { "epoch": 0.5398804223381249, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.119854927062988, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8623088598251343, "num_tokens": 162067130.0, "step": 4244 }, { "epoch": 0.5400076326167155, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.112343788146973, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8601551055908203, "num_tokens": 162107452.0, "step": 4245 }, { "epoch": 0.5401348428953059, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.153788566589355, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8580528497695923, "num_tokens": 162144220.0, "step": 4246 }, { "epoch": 0.5402620531738964, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.166163444519043, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8679684400558472, "num_tokens": 162181798.0, "step": 4247 }, { "epoch": 0.540389263452487, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.03859806060791, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8513018488883972, "num_tokens": 162222501.0, "step": 4248 }, { "epoch": 0.5405164737310775, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.11313247680664, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8553297519683838, "num_tokens": 162265533.0, "step": 4249 }, { "epoch": 0.540643684009668, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.128536224365234, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8735854029655457, "num_tokens": 162304984.0, "step": 4250 }, { "epoch": 0.5407708942882585, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.075636863708496, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8499429821968079, "num_tokens": 162342358.0, "step": 4251 }, { "epoch": 0.540898104566849, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.174814224243164, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8603166937828064, "num_tokens": 162377343.0, "step": 4252 }, { "epoch": 0.5410253148454395, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.093122482299805, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8592541813850403, "num_tokens": 162416416.0, "step": 4253 }, { "epoch": 0.54115252512403, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.105414390563965, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8846049308776855, "num_tokens": 162447108.0, "step": 4254 }, { "epoch": 0.5412797354026205, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.121471405029297, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8609782457351685, "num_tokens": 162489659.0, "step": 4255 }, { "epoch": 0.5414069456812111, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.099091529846191, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.8491733074188232, "num_tokens": 162525816.0, "step": 4256 }, { "epoch": 0.5415341559598016, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.16515827178955, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.854781985282898, "num_tokens": 162564154.0, "step": 4257 }, { "epoch": 0.5416613662383921, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.0524263381958, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8529640436172485, "num_tokens": 162602728.0, "step": 4258 }, { "epoch": 0.5417885765169825, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.146088600158691, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8637804985046387, "num_tokens": 162636659.0, "step": 4259 }, { "epoch": 0.5419157867955731, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.15440559387207, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8749228715896606, "num_tokens": 162676639.0, "step": 4260 }, { "epoch": 0.5420429970741636, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.105428695678711, "learning_rate": 1e-06, "loss": 0.4957, "mean_token_accuracy": 0.8409175872802734, "num_tokens": 162716763.0, "step": 4261 }, { "epoch": 0.5421702073527541, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.158089637756348, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8480832576751709, "num_tokens": 162753943.0, "step": 4262 }, { "epoch": 0.5422974176313446, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.100363731384277, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8539175391197205, "num_tokens": 162792938.0, "step": 4263 }, { "epoch": 0.5424246279099352, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.100082397460938, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8566530346870422, "num_tokens": 162838938.0, "step": 4264 }, { "epoch": 0.5425518381885256, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.144902229309082, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8645422458648682, "num_tokens": 162875026.0, "step": 4265 }, { "epoch": 0.5426790484671161, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.129436492919922, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8614310622215271, "num_tokens": 162910724.0, "step": 4266 }, { "epoch": 0.5428062587457066, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.133390426635742, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8623093962669373, "num_tokens": 162941602.0, "step": 4267 }, { "epoch": 0.5429334690242972, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.136258125305176, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8458917737007141, "num_tokens": 162980760.0, "step": 4268 }, { "epoch": 0.5430606793028877, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.075967788696289, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8685703277587891, "num_tokens": 163016155.0, "step": 4269 }, { "epoch": 0.5431878895814782, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.158026695251465, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8499194383621216, "num_tokens": 163052519.0, "step": 4270 }, { "epoch": 0.5433150998600687, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.167317390441895, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8675860166549683, "num_tokens": 163091806.0, "step": 4271 }, { "epoch": 0.5434423101386592, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.212044715881348, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8736447691917419, "num_tokens": 163128830.0, "step": 4272 }, { "epoch": 0.5435695204172497, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.147380828857422, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8537907600402832, "num_tokens": 163169567.0, "step": 4273 }, { "epoch": 0.5436967306958402, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.220988273620605, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8517812490463257, "num_tokens": 163202913.0, "step": 4274 }, { "epoch": 0.5438239409744308, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.182220458984375, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8676198720932007, "num_tokens": 163236741.0, "step": 4275 }, { "epoch": 0.5439511512530213, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.2435884475708, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8528912663459778, "num_tokens": 163281402.0, "step": 4276 }, { "epoch": 0.5440783615316117, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.126317977905273, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.868064820766449, "num_tokens": 163325288.0, "step": 4277 }, { "epoch": 0.5442055718102022, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.152356147766113, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8486617803573608, "num_tokens": 163367471.0, "step": 4278 }, { "epoch": 0.5443327820887928, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.233270645141602, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8603212237358093, "num_tokens": 163400077.0, "step": 4279 }, { "epoch": 0.5444599923673833, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.213199615478516, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.862273633480072, "num_tokens": 163433739.0, "step": 4280 }, { "epoch": 0.5445872026459738, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.2210111618042, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8591651916503906, "num_tokens": 163467447.0, "step": 4281 }, { "epoch": 0.5447144129245644, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.165947914123535, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8565409183502197, "num_tokens": 163506633.0, "step": 4282 }, { "epoch": 0.5448416232031548, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.214922904968262, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.8443673849105835, "num_tokens": 163542773.0, "step": 4283 }, { "epoch": 0.5449688334817453, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.166242599487305, "learning_rate": 1e-06, "loss": 0.5412, "mean_token_accuracy": 0.8260772228240967, "num_tokens": 163585247.0, "step": 4284 }, { "epoch": 0.5450960437603358, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.220505714416504, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.847517728805542, "num_tokens": 163622802.0, "step": 4285 }, { "epoch": 0.5452232540389264, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.203235626220703, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8830080032348633, "num_tokens": 163655980.0, "step": 4286 }, { "epoch": 0.5453504643175169, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.226384162902832, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8643323183059692, "num_tokens": 163691494.0, "step": 4287 }, { "epoch": 0.5454776745961074, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.163729667663574, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.870003342628479, "num_tokens": 163725169.0, "step": 4288 }, { "epoch": 0.5456048848746978, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.188244819641113, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8628681898117065, "num_tokens": 163766050.0, "step": 4289 }, { "epoch": 0.5457320951532884, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.18620491027832, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8535158038139343, "num_tokens": 163799559.0, "step": 4290 }, { "epoch": 0.5458593054318789, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.205750465393066, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8539657592773438, "num_tokens": 163837052.0, "step": 4291 }, { "epoch": 0.5459865157104694, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.217391967773438, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.856494665145874, "num_tokens": 163872494.0, "step": 4292 }, { "epoch": 0.5461137259890599, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.255465507507324, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8514349460601807, "num_tokens": 163913499.0, "step": 4293 }, { "epoch": 0.5462409362676505, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.16008186340332, "learning_rate": 1e-06, "loss": 0.4961, "mean_token_accuracy": 0.8403722047805786, "num_tokens": 163957965.0, "step": 4294 }, { "epoch": 0.5463681465462409, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.201000213623047, "learning_rate": 1e-06, "loss": 0.5681, "mean_token_accuracy": 0.8182979822158813, "num_tokens": 164005661.0, "step": 4295 }, { "epoch": 0.5464953568248314, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.188549995422363, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8744269013404846, "num_tokens": 164044717.0, "step": 4296 }, { "epoch": 0.5466225671034219, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.215715408325195, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8583990335464478, "num_tokens": 164080022.0, "step": 4297 }, { "epoch": 0.5467497773820125, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.289166450500488, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.866185188293457, "num_tokens": 164112749.0, "step": 4298 }, { "epoch": 0.546876987660603, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.190486907958984, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8530302047729492, "num_tokens": 164152517.0, "step": 4299 }, { "epoch": 0.5470041979391935, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.150862693786621, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8652827739715576, "num_tokens": 164195245.0, "step": 4300 }, { "epoch": 0.5471314082177839, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.21849536895752, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8540472984313965, "num_tokens": 164230643.0, "step": 4301 }, { "epoch": 0.5472586184963745, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.195096015930176, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8558775186538696, "num_tokens": 164271089.0, "step": 4302 }, { "epoch": 0.547385828774965, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.134442329406738, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8568935990333557, "num_tokens": 164310618.0, "step": 4303 }, { "epoch": 0.5475130390535555, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.213054656982422, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8451781272888184, "num_tokens": 164348630.0, "step": 4304 }, { "epoch": 0.5476402493321461, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.189213752746582, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8656593561172485, "num_tokens": 164383735.0, "step": 4305 }, { "epoch": 0.5477674596107366, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.162521362304688, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8602486848831177, "num_tokens": 164426115.0, "step": 4306 }, { "epoch": 0.5478946698893271, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.144806861877441, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8511568307876587, "num_tokens": 164462119.0, "step": 4307 }, { "epoch": 0.5480218801679175, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.220076560974121, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.855721652507782, "num_tokens": 164503030.0, "step": 4308 }, { "epoch": 0.5481490904465081, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.166911125183105, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8613645434379578, "num_tokens": 164541091.0, "step": 4309 }, { "epoch": 0.5482763007250986, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.172892570495605, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8530704975128174, "num_tokens": 164577678.0, "step": 4310 }, { "epoch": 0.5484035110036891, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.155837059020996, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8561866879463196, "num_tokens": 164612779.0, "step": 4311 }, { "epoch": 0.5485307212822796, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.179262161254883, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8643922805786133, "num_tokens": 164649849.0, "step": 4312 }, { "epoch": 0.5486579315608702, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.190735816955566, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8559209704399109, "num_tokens": 164690538.0, "step": 4313 }, { "epoch": 0.5487851418394606, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.192617416381836, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8590598106384277, "num_tokens": 164729338.0, "step": 4314 }, { "epoch": 0.5489123521180511, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.186786651611328, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8702094554901123, "num_tokens": 164766648.0, "step": 4315 }, { "epoch": 0.5490395623966416, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.23857307434082, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8580246567726135, "num_tokens": 164804254.0, "step": 4316 }, { "epoch": 0.5491667726752322, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.123306274414062, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8712251782417297, "num_tokens": 164839943.0, "step": 4317 }, { "epoch": 0.5492939829538227, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.179314613342285, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8507255911827087, "num_tokens": 164887057.0, "step": 4318 }, { "epoch": 0.5494211932324132, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.188427925109863, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8583918809890747, "num_tokens": 164928515.0, "step": 4319 }, { "epoch": 0.5495484035110036, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.164898872375488, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8580597043037415, "num_tokens": 164963818.0, "step": 4320 }, { "epoch": 0.5496756137895942, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.211581230163574, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8722175359725952, "num_tokens": 165001091.0, "step": 4321 }, { "epoch": 0.5498028240681847, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.185285568237305, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8416023254394531, "num_tokens": 165046009.0, "step": 4322 }, { "epoch": 0.5499300343467752, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.13925552368164, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8556181788444519, "num_tokens": 165090729.0, "step": 4323 }, { "epoch": 0.5500572446253658, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.219361305236816, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8756197690963745, "num_tokens": 165129771.0, "step": 4324 }, { "epoch": 0.5501844549039563, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.191941261291504, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8734802603721619, "num_tokens": 165164310.0, "step": 4325 }, { "epoch": 0.5503116651825467, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.172103881835938, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.855599045753479, "num_tokens": 165202842.0, "step": 4326 }, { "epoch": 0.5504388754611372, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.19150161743164, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8627064228057861, "num_tokens": 165235998.0, "step": 4327 }, { "epoch": 0.5505660857397278, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.192022323608398, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8601061105728149, "num_tokens": 165271779.0, "step": 4328 }, { "epoch": 0.5506932960183183, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.214077949523926, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.853092610836029, "num_tokens": 165306136.0, "step": 4329 }, { "epoch": 0.5508205062969088, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.192156791687012, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8737771511077881, "num_tokens": 165342401.0, "step": 4330 }, { "epoch": 0.5509477165754993, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.132830619812012, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8617153167724609, "num_tokens": 165380120.0, "step": 4331 }, { "epoch": 0.5510749268540898, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.175904273986816, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8461048603057861, "num_tokens": 165418210.0, "step": 4332 }, { "epoch": 0.5512021371326803, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.197396278381348, "learning_rate": 1e-06, "loss": 0.51, "mean_token_accuracy": 0.8382195830345154, "num_tokens": 165460952.0, "step": 4333 }, { "epoch": 0.5513293474112708, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.184745788574219, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8717460632324219, "num_tokens": 165500370.0, "step": 4334 }, { "epoch": 0.5514565576898613, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.222105026245117, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8588424921035767, "num_tokens": 165534089.0, "step": 4335 }, { "epoch": 0.5515837679684519, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.184274673461914, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8514294028282166, "num_tokens": 165573111.0, "step": 4336 }, { "epoch": 0.5517109782470424, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.171170234680176, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8487828969955444, "num_tokens": 165611942.0, "step": 4337 }, { "epoch": 0.5518381885256328, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.195802688598633, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.865791916847229, "num_tokens": 165651703.0, "step": 4338 }, { "epoch": 0.5519653988042234, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.108855247497559, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8566617965698242, "num_tokens": 165691257.0, "step": 4339 }, { "epoch": 0.5520926090828139, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.211125373840332, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8748924732208252, "num_tokens": 165727947.0, "step": 4340 }, { "epoch": 0.5522198193614044, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.188139915466309, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8679710626602173, "num_tokens": 165763400.0, "step": 4341 }, { "epoch": 0.5523470296399949, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.143314361572266, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.858869194984436, "num_tokens": 165809586.0, "step": 4342 }, { "epoch": 0.5524742399185855, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.224871635437012, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8612623810768127, "num_tokens": 165839565.0, "step": 4343 }, { "epoch": 0.5526014501971759, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.228821754455566, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8627108931541443, "num_tokens": 165873539.0, "step": 4344 }, { "epoch": 0.5527286604757664, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.111130714416504, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8676387071609497, "num_tokens": 165920987.0, "step": 4345 }, { "epoch": 0.5528558707543569, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.196261405944824, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8482614755630493, "num_tokens": 165965661.0, "step": 4346 }, { "epoch": 0.5529830810329475, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.161931991577148, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8527316451072693, "num_tokens": 165998239.0, "step": 4347 }, { "epoch": 0.553110291311538, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.192818641662598, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.856681227684021, "num_tokens": 166036489.0, "step": 4348 }, { "epoch": 0.5532375015901285, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.19291877746582, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8549249172210693, "num_tokens": 166078774.0, "step": 4349 }, { "epoch": 0.5533647118687189, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.125020027160645, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8657528162002563, "num_tokens": 166119555.0, "step": 4350 }, { "epoch": 0.5534919221473095, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.195691108703613, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8573458194732666, "num_tokens": 166159788.0, "step": 4351 }, { "epoch": 0.5536191324259, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.178350448608398, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8536465764045715, "num_tokens": 166200230.0, "step": 4352 }, { "epoch": 0.5537463427044905, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.108269691467285, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.862830400466919, "num_tokens": 166237864.0, "step": 4353 }, { "epoch": 0.553873552983081, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.228066444396973, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8612624406814575, "num_tokens": 166278347.0, "step": 4354 }, { "epoch": 0.5540007632616716, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.204422950744629, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8505525588989258, "num_tokens": 166315744.0, "step": 4355 }, { "epoch": 0.554127973540262, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.145841598510742, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8598663806915283, "num_tokens": 166352705.0, "step": 4356 }, { "epoch": 0.5542551838188525, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.22746753692627, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.863697350025177, "num_tokens": 166393956.0, "step": 4357 }, { "epoch": 0.554382394097443, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.134126663208008, "learning_rate": 1e-06, "loss": 0.5198, "mean_token_accuracy": 0.836556077003479, "num_tokens": 166430172.0, "step": 4358 }, { "epoch": 0.5545096043760336, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.218057632446289, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8589686751365662, "num_tokens": 166468732.0, "step": 4359 }, { "epoch": 0.5546368146546241, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.145292282104492, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8551643490791321, "num_tokens": 166511383.0, "step": 4360 }, { "epoch": 0.5547640249332146, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.152771949768066, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8629618883132935, "num_tokens": 166547674.0, "step": 4361 }, { "epoch": 0.5548912352118052, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.138273239135742, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.851495623588562, "num_tokens": 166581763.0, "step": 4362 }, { "epoch": 0.5550184454903956, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.240525245666504, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8563823699951172, "num_tokens": 166618604.0, "step": 4363 }, { "epoch": 0.5551456557689861, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.197640419006348, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8682801127433777, "num_tokens": 166654812.0, "step": 4364 }, { "epoch": 0.5552728660475766, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.215882301330566, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.845244288444519, "num_tokens": 166693206.0, "step": 4365 }, { "epoch": 0.5554000763261672, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.189549446105957, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8577773571014404, "num_tokens": 166734531.0, "step": 4366 }, { "epoch": 0.5555272866047577, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.230981826782227, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8567761182785034, "num_tokens": 166771362.0, "step": 4367 }, { "epoch": 0.5556544968833482, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.231660842895508, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8496950268745422, "num_tokens": 166802784.0, "step": 4368 }, { "epoch": 0.5557817071619386, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.237351417541504, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8547623753547668, "num_tokens": 166835590.0, "step": 4369 }, { "epoch": 0.5559089174405292, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.254186630249023, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8530494570732117, "num_tokens": 166872145.0, "step": 4370 }, { "epoch": 0.5560361277191197, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.173075675964355, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8544655442237854, "num_tokens": 166912855.0, "step": 4371 }, { "epoch": 0.5561633379977102, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.20566177368164, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8434206247329712, "num_tokens": 166949482.0, "step": 4372 }, { "epoch": 0.5562905482763008, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.20529556274414, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8589445352554321, "num_tokens": 166986727.0, "step": 4373 }, { "epoch": 0.5564177585548913, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.223851203918457, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8515525460243225, "num_tokens": 167025869.0, "step": 4374 }, { "epoch": 0.5565449688334817, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.271100997924805, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8687223196029663, "num_tokens": 167059825.0, "step": 4375 }, { "epoch": 0.5566721791120722, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.14306640625, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8630884885787964, "num_tokens": 167097375.0, "step": 4376 }, { "epoch": 0.5567993893906628, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.18941879272461, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8494272828102112, "num_tokens": 167136555.0, "step": 4377 }, { "epoch": 0.5569265996692533, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.140856742858887, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.85627281665802, "num_tokens": 167179870.0, "step": 4378 }, { "epoch": 0.5570538099478438, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.1821928024292, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.857263445854187, "num_tokens": 167215242.0, "step": 4379 }, { "epoch": 0.5571810202264343, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.225152015686035, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8655262589454651, "num_tokens": 167250690.0, "step": 4380 }, { "epoch": 0.5573082305050248, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.143463134765625, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.845382571220398, "num_tokens": 167283677.0, "step": 4381 }, { "epoch": 0.5574354407836153, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.238863945007324, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8593825697898865, "num_tokens": 167321355.0, "step": 4382 }, { "epoch": 0.5575626510622058, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.162588119506836, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8511173725128174, "num_tokens": 167363103.0, "step": 4383 }, { "epoch": 0.5576898613407963, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.153413772583008, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8718011379241943, "num_tokens": 167399732.0, "step": 4384 }, { "epoch": 0.5578170716193869, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.211275100708008, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.85710608959198, "num_tokens": 167434938.0, "step": 4385 }, { "epoch": 0.5579442818979774, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.186291694641113, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8618611693382263, "num_tokens": 167472293.0, "step": 4386 }, { "epoch": 0.5580714921765678, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.168947219848633, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.863670825958252, "num_tokens": 167511015.0, "step": 4387 }, { "epoch": 0.5581987024551583, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.207411766052246, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8570493459701538, "num_tokens": 167549777.0, "step": 4388 }, { "epoch": 0.5583259127337489, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.197391510009766, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8644417524337769, "num_tokens": 167592342.0, "step": 4389 }, { "epoch": 0.5584531230123394, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.11720085144043, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8620498180389404, "num_tokens": 167630829.0, "step": 4390 }, { "epoch": 0.5585803332909299, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.280218124389648, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8471958041191101, "num_tokens": 167675938.0, "step": 4391 }, { "epoch": 0.5587075435695205, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.110445022583008, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8623080253601074, "num_tokens": 167716550.0, "step": 4392 }, { "epoch": 0.5588347538481109, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.19558048248291, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8457550406455994, "num_tokens": 167754401.0, "step": 4393 }, { "epoch": 0.5589619641267014, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.208806991577148, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.862324595451355, "num_tokens": 167792124.0, "step": 4394 }, { "epoch": 0.5590891744052919, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.183418273925781, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8767558932304382, "num_tokens": 167829226.0, "step": 4395 }, { "epoch": 0.5592163846838825, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.207956314086914, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.868975818157196, "num_tokens": 167863439.0, "step": 4396 }, { "epoch": 0.559343594962473, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.132709503173828, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8590791821479797, "num_tokens": 167907019.0, "step": 4397 }, { "epoch": 0.5594708052410635, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.208120346069336, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8503754734992981, "num_tokens": 167942539.0, "step": 4398 }, { "epoch": 0.5595980155196539, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.288362503051758, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8581123352050781, "num_tokens": 167982848.0, "step": 4399 }, { "epoch": 0.5597252257982445, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.252298355102539, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8673275113105774, "num_tokens": 168017500.0, "step": 4400 }, { "epoch": 0.559852436076835, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.225260734558105, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8434056043624878, "num_tokens": 168052790.0, "step": 4401 }, { "epoch": 0.5599796463554255, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.272754669189453, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8660595417022705, "num_tokens": 168082541.0, "step": 4402 }, { "epoch": 0.560106856634016, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.25744342803955, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8636507987976074, "num_tokens": 168128218.0, "step": 4403 }, { "epoch": 0.5602340669126066, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.213521957397461, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8612790107727051, "num_tokens": 168165496.0, "step": 4404 }, { "epoch": 0.560361277191197, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.260759353637695, "learning_rate": 1e-06, "loss": 0.4881, "mean_token_accuracy": 0.8453897833824158, "num_tokens": 168204675.0, "step": 4405 }, { "epoch": 0.5604884874697875, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.216056823730469, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8547500371932983, "num_tokens": 168244832.0, "step": 4406 }, { "epoch": 0.560615697748378, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.294891357421875, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8539301156997681, "num_tokens": 168282440.0, "step": 4407 }, { "epoch": 0.5607429080269686, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.168390274047852, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8644577264785767, "num_tokens": 168325303.0, "step": 4408 }, { "epoch": 0.5608701183055591, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.283629417419434, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8649643063545227, "num_tokens": 168360628.0, "step": 4409 }, { "epoch": 0.5609973285841496, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.230125427246094, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8463022112846375, "num_tokens": 168398601.0, "step": 4410 }, { "epoch": 0.5611245388627402, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.284370422363281, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.844717264175415, "num_tokens": 168436893.0, "step": 4411 }, { "epoch": 0.5612517491413306, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.168731689453125, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8547270894050598, "num_tokens": 168478794.0, "step": 4412 }, { "epoch": 0.5613789594199211, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.228742599487305, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.855934202671051, "num_tokens": 168520184.0, "step": 4413 }, { "epoch": 0.5615061696985116, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.22507381439209, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8557378649711609, "num_tokens": 168554305.0, "step": 4414 }, { "epoch": 0.5616333799771022, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.210065841674805, "learning_rate": 1e-06, "loss": 0.493, "mean_token_accuracy": 0.8425616025924683, "num_tokens": 168597752.0, "step": 4415 }, { "epoch": 0.5617605902556927, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.209547996520996, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8641741871833801, "num_tokens": 168637764.0, "step": 4416 }, { "epoch": 0.5618878005342832, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.280292510986328, "learning_rate": 1e-06, "loss": 0.4875, "mean_token_accuracy": 0.8391417264938354, "num_tokens": 168672260.0, "step": 4417 }, { "epoch": 0.5620150108128736, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.231953620910645, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8671900629997253, "num_tokens": 168711478.0, "step": 4418 }, { "epoch": 0.5621422210914642, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.22311019897461, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8612995147705078, "num_tokens": 168751773.0, "step": 4419 }, { "epoch": 0.5622694313700547, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.231637954711914, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8687787055969238, "num_tokens": 168786364.0, "step": 4420 }, { "epoch": 0.5623966416486452, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.258322715759277, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8555089235305786, "num_tokens": 168826819.0, "step": 4421 }, { "epoch": 0.5625238519272358, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.168107032775879, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8566004037857056, "num_tokens": 168869405.0, "step": 4422 }, { "epoch": 0.5626510622058263, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.282154083251953, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.864558756351471, "num_tokens": 168904323.0, "step": 4423 }, { "epoch": 0.5627782724844167, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.18066120147705, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.849115788936615, "num_tokens": 168948654.0, "step": 4424 }, { "epoch": 0.5629054827630072, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.252397537231445, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8646609783172607, "num_tokens": 168984333.0, "step": 4425 }, { "epoch": 0.5630326930415978, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.267528533935547, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.859550952911377, "num_tokens": 169019843.0, "step": 4426 }, { "epoch": 0.5631599033201883, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.234688758850098, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8691663146018982, "num_tokens": 169058177.0, "step": 4427 }, { "epoch": 0.5632871135987788, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.200549125671387, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8637998104095459, "num_tokens": 169102268.0, "step": 4428 }, { "epoch": 0.5634143238773693, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.169816970825195, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8539990186691284, "num_tokens": 169141709.0, "step": 4429 }, { "epoch": 0.5635415341559598, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.248541831970215, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8554438948631287, "num_tokens": 169177107.0, "step": 4430 }, { "epoch": 0.5636687444345503, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.22048568725586, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.855510950088501, "num_tokens": 169211531.0, "step": 4431 }, { "epoch": 0.5637959547131408, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.233675956726074, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8513233065605164, "num_tokens": 169247507.0, "step": 4432 }, { "epoch": 0.5639231649917313, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.240287780761719, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8722145557403564, "num_tokens": 169281007.0, "step": 4433 }, { "epoch": 0.5640503752703219, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.245512008666992, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8598870038986206, "num_tokens": 169322295.0, "step": 4434 }, { "epoch": 0.5641775855489124, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.23507308959961, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8554233908653259, "num_tokens": 169358184.0, "step": 4435 }, { "epoch": 0.5643047958275028, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.211989402770996, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8639322519302368, "num_tokens": 169391518.0, "step": 4436 }, { "epoch": 0.5644320061060933, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.25607681274414, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8586660623550415, "num_tokens": 169426743.0, "step": 4437 }, { "epoch": 0.5645592163846839, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.186963081359863, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8774847984313965, "num_tokens": 169466452.0, "step": 4438 }, { "epoch": 0.5646864266632744, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.281728744506836, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.863959550857544, "num_tokens": 169500295.0, "step": 4439 }, { "epoch": 0.5648136369418649, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.198661804199219, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8581183552742004, "num_tokens": 169541829.0, "step": 4440 }, { "epoch": 0.5649408472204555, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.167744636535645, "learning_rate": 1e-06, "loss": 0.4886, "mean_token_accuracy": 0.8435800671577454, "num_tokens": 169581547.0, "step": 4441 }, { "epoch": 0.5650680574990459, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.166396141052246, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.863689661026001, "num_tokens": 169622107.0, "step": 4442 }, { "epoch": 0.5651952677776364, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.176573753356934, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8550693392753601, "num_tokens": 169657258.0, "step": 4443 }, { "epoch": 0.5653224780562269, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.222213745117188, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8478776812553406, "num_tokens": 169691568.0, "step": 4444 }, { "epoch": 0.5654496883348175, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.228120803833008, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8653090596199036, "num_tokens": 169732429.0, "step": 4445 }, { "epoch": 0.565576898613408, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.169953346252441, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8519623279571533, "num_tokens": 169769224.0, "step": 4446 }, { "epoch": 0.5657041088919985, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.225326538085938, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8582335710525513, "num_tokens": 169812830.0, "step": 4447 }, { "epoch": 0.5658313191705889, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.2136869430542, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.855879545211792, "num_tokens": 169843823.0, "step": 4448 }, { "epoch": 0.5659585294491795, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.227821350097656, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8604625463485718, "num_tokens": 169882847.0, "step": 4449 }, { "epoch": 0.56608573972777, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.193670272827148, "learning_rate": 1e-06, "loss": 0.4927, "mean_token_accuracy": 0.8403753042221069, "num_tokens": 169921264.0, "step": 4450 }, { "epoch": 0.5662129500063605, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.270402908325195, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8652424812316895, "num_tokens": 169949681.0, "step": 4451 }, { "epoch": 0.566340160284951, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.186811447143555, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8727630972862244, "num_tokens": 169986254.0, "step": 4452 }, { "epoch": 0.5664673705635416, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.230758666992188, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8551561236381531, "num_tokens": 170022745.0, "step": 4453 }, { "epoch": 0.566594580842132, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.213273048400879, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.869378387928009, "num_tokens": 170057038.0, "step": 4454 }, { "epoch": 0.5667217911207225, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.192453384399414, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8773332238197327, "num_tokens": 170098460.0, "step": 4455 }, { "epoch": 0.566849001399313, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.2050199508667, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.846526563167572, "num_tokens": 170137689.0, "step": 4456 }, { "epoch": 0.5669762116779036, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.16064167022705, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8599661588668823, "num_tokens": 170176716.0, "step": 4457 }, { "epoch": 0.5671034219564941, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.251969337463379, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8556036949157715, "num_tokens": 170212621.0, "step": 4458 }, { "epoch": 0.5672306322350846, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.145524024963379, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8489780426025391, "num_tokens": 170254493.0, "step": 4459 }, { "epoch": 0.5673578425136752, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.238972663879395, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8801068663597107, "num_tokens": 170290559.0, "step": 4460 }, { "epoch": 0.5674850527922656, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.157195091247559, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.852489173412323, "num_tokens": 170335079.0, "step": 4461 }, { "epoch": 0.5676122630708561, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.232536315917969, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8621780872344971, "num_tokens": 170370665.0, "step": 4462 }, { "epoch": 0.5677394733494466, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.222049713134766, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8583167791366577, "num_tokens": 170412585.0, "step": 4463 }, { "epoch": 0.5678666836280372, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.221847534179688, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8655211925506592, "num_tokens": 170450806.0, "step": 4464 }, { "epoch": 0.5679938939066277, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.257880210876465, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8592059016227722, "num_tokens": 170483265.0, "step": 4465 }, { "epoch": 0.5681211041852182, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.266538619995117, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8750553131103516, "num_tokens": 170520910.0, "step": 4466 }, { "epoch": 0.5682483144638086, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.252646446228027, "learning_rate": 1e-06, "loss": 0.4906, "mean_token_accuracy": 0.8420381546020508, "num_tokens": 170556516.0, "step": 4467 }, { "epoch": 0.5683755247423992, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.28607177734375, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8569365739822388, "num_tokens": 170595571.0, "step": 4468 }, { "epoch": 0.5685027350209897, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.201766014099121, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8786551356315613, "num_tokens": 170634117.0, "step": 4469 }, { "epoch": 0.5686299452995802, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.203021049499512, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8539509773254395, "num_tokens": 170677624.0, "step": 4470 }, { "epoch": 0.5687571555781707, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.267938613891602, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8440811634063721, "num_tokens": 170716732.0, "step": 4471 }, { "epoch": 0.5688843658567613, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.175576210021973, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8522857427597046, "num_tokens": 170758521.0, "step": 4472 }, { "epoch": 0.5690115761353517, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.186107635498047, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8582370281219482, "num_tokens": 170796799.0, "step": 4473 }, { "epoch": 0.5691387864139422, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.251267433166504, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8447360992431641, "num_tokens": 170838569.0, "step": 4474 }, { "epoch": 0.5692659966925327, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.18014907836914, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8609554767608643, "num_tokens": 170878206.0, "step": 4475 }, { "epoch": 0.5693932069711233, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.249200820922852, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8465816974639893, "num_tokens": 170915506.0, "step": 4476 }, { "epoch": 0.5695204172497138, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.2215576171875, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8554756045341492, "num_tokens": 170952303.0, "step": 4477 }, { "epoch": 0.5696476275283043, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.23005199432373, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.85172039270401, "num_tokens": 170993757.0, "step": 4478 }, { "epoch": 0.5697748378068948, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.232366561889648, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8567320108413696, "num_tokens": 171030240.0, "step": 4479 }, { "epoch": 0.5699020480854853, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.30941104888916, "learning_rate": 1e-06, "loss": 0.5031, "mean_token_accuracy": 0.8354288339614868, "num_tokens": 171069654.0, "step": 4480 }, { "epoch": 0.5700292583640758, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.225462913513184, "learning_rate": 1e-06, "loss": 0.4905, "mean_token_accuracy": 0.8422775864601135, "num_tokens": 171107341.0, "step": 4481 }, { "epoch": 0.5701564686426663, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.212428092956543, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8498990535736084, "num_tokens": 171144063.0, "step": 4482 }, { "epoch": 0.5702836789212569, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.285246849060059, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8532178401947021, "num_tokens": 171180448.0, "step": 4483 }, { "epoch": 0.5704108891998474, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.239943504333496, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8627458810806274, "num_tokens": 171214220.0, "step": 4484 }, { "epoch": 0.5705380994784378, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.170636177062988, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8645704984664917, "num_tokens": 171252798.0, "step": 4485 }, { "epoch": 0.5706653097570283, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.214742660522461, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8639870882034302, "num_tokens": 171295434.0, "step": 4486 }, { "epoch": 0.5707925200356189, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.21740436553955, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8602064847946167, "num_tokens": 171334129.0, "step": 4487 }, { "epoch": 0.5709197303142094, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.298712730407715, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8542869091033936, "num_tokens": 171370239.0, "step": 4488 }, { "epoch": 0.5710469405927999, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.2796630859375, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8730757236480713, "num_tokens": 171402411.0, "step": 4489 }, { "epoch": 0.5711741508713905, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.260711669921875, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8686228394508362, "num_tokens": 171438736.0, "step": 4490 }, { "epoch": 0.5713013611499809, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.262743949890137, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8649755716323853, "num_tokens": 171477949.0, "step": 4491 }, { "epoch": 0.5714285714285714, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.331582069396973, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8537602424621582, "num_tokens": 171520953.0, "step": 4492 }, { "epoch": 0.5715557817071619, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.218605995178223, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8649666905403137, "num_tokens": 171559415.0, "step": 4493 }, { "epoch": 0.5716829919857525, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.306336402893066, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8663438558578491, "num_tokens": 171593118.0, "step": 4494 }, { "epoch": 0.571810202264343, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.273640632629395, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8459730744361877, "num_tokens": 171634514.0, "step": 4495 }, { "epoch": 0.5719374125429335, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.255616188049316, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8622220158576965, "num_tokens": 171675239.0, "step": 4496 }, { "epoch": 0.5720646228215239, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.209843635559082, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8660043478012085, "num_tokens": 171713154.0, "step": 4497 }, { "epoch": 0.5721918331001145, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.228073120117188, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8521610498428345, "num_tokens": 171750046.0, "step": 4498 }, { "epoch": 0.572319043378705, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.311175346374512, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8483865261077881, "num_tokens": 171782708.0, "step": 4499 }, { "epoch": 0.5724462536572955, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.214090347290039, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8545360565185547, "num_tokens": 171819421.0, "step": 4500 }, { "epoch": 0.572573463935886, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.269927978515625, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8571946620941162, "num_tokens": 171860993.0, "step": 4501 }, { "epoch": 0.5727006742144766, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.241995811462402, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8581250905990601, "num_tokens": 171896894.0, "step": 4502 }, { "epoch": 0.572827884493067, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.286881446838379, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8561424016952515, "num_tokens": 171938694.0, "step": 4503 }, { "epoch": 0.5729550947716575, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.171846389770508, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8827070593833923, "num_tokens": 171972606.0, "step": 4504 }, { "epoch": 0.573082305050248, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.237293243408203, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.860584020614624, "num_tokens": 172010861.0, "step": 4505 }, { "epoch": 0.5732095153288386, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.277238845825195, "learning_rate": 1e-06, "loss": 0.4947, "mean_token_accuracy": 0.8438736200332642, "num_tokens": 172049867.0, "step": 4506 }, { "epoch": 0.5733367256074291, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.18795108795166, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.859768271446228, "num_tokens": 172092176.0, "step": 4507 }, { "epoch": 0.5734639358860196, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.263551712036133, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8584185838699341, "num_tokens": 172132789.0, "step": 4508 }, { "epoch": 0.5735911461646102, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.192344665527344, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8730001449584961, "num_tokens": 172170536.0, "step": 4509 }, { "epoch": 0.5737183564432006, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.220763206481934, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8581308126449585, "num_tokens": 172206828.0, "step": 4510 }, { "epoch": 0.5738455667217911, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.248516082763672, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8498883247375488, "num_tokens": 172242311.0, "step": 4511 }, { "epoch": 0.5739727770003816, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.278285026550293, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8535770177841187, "num_tokens": 172275135.0, "step": 4512 }, { "epoch": 0.5740999872789722, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.195270538330078, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8669941425323486, "num_tokens": 172308682.0, "step": 4513 }, { "epoch": 0.5742271975575627, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.226618766784668, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8532202839851379, "num_tokens": 172345834.0, "step": 4514 }, { "epoch": 0.5743544078361532, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.148499488830566, "learning_rate": 1e-06, "loss": 0.5202, "mean_token_accuracy": 0.8360228538513184, "num_tokens": 172387976.0, "step": 4515 }, { "epoch": 0.5744816181147436, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.269027709960938, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8577426671981812, "num_tokens": 172427447.0, "step": 4516 }, { "epoch": 0.5746088283933342, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.26021957397461, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.849799394607544, "num_tokens": 172468686.0, "step": 4517 }, { "epoch": 0.5747360386719247, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.27786636352539, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8541985750198364, "num_tokens": 172508190.0, "step": 4518 }, { "epoch": 0.5748632489505152, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.222326278686523, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8740594387054443, "num_tokens": 172552295.0, "step": 4519 }, { "epoch": 0.5749904592291057, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.2937593460083, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8633872270584106, "num_tokens": 172591065.0, "step": 4520 }, { "epoch": 0.5751176695076963, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.265772819519043, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.86106276512146, "num_tokens": 172633741.0, "step": 4521 }, { "epoch": 0.5752448797862867, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.265341758728027, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8680810928344727, "num_tokens": 172669081.0, "step": 4522 }, { "epoch": 0.5753720900648772, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.267736434936523, "learning_rate": 1e-06, "loss": 0.5148, "mean_token_accuracy": 0.8364667892456055, "num_tokens": 172704956.0, "step": 4523 }, { "epoch": 0.5754993003434677, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.339369773864746, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8625035285949707, "num_tokens": 172742270.0, "step": 4524 }, { "epoch": 0.5756265106220583, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.286131858825684, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8468641042709351, "num_tokens": 172778065.0, "step": 4525 }, { "epoch": 0.5757537209006488, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.217286109924316, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8667334318161011, "num_tokens": 172815825.0, "step": 4526 }, { "epoch": 0.5758809311792393, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.269479751586914, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8585780262947083, "num_tokens": 172849560.0, "step": 4527 }, { "epoch": 0.5760081414578297, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.277766227722168, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8597097396850586, "num_tokens": 172885203.0, "step": 4528 }, { "epoch": 0.5761353517364203, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.24207878112793, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8577249646186829, "num_tokens": 172922341.0, "step": 4529 }, { "epoch": 0.5762625620150108, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.361120223999023, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8641878366470337, "num_tokens": 172953363.0, "step": 4530 }, { "epoch": 0.5763897722936013, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.28434944152832, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8636445999145508, "num_tokens": 172986253.0, "step": 4531 }, { "epoch": 0.5765169825721919, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.326639175415039, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8695964813232422, "num_tokens": 173023466.0, "step": 4532 }, { "epoch": 0.5766441928507824, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.274405479431152, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8530789613723755, "num_tokens": 173063329.0, "step": 4533 }, { "epoch": 0.5767714031293728, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.300105094909668, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.864425778388977, "num_tokens": 173100676.0, "step": 4534 }, { "epoch": 0.5768986134079633, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.255090713500977, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8691564202308655, "num_tokens": 173135088.0, "step": 4535 }, { "epoch": 0.5770258236865539, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.262482643127441, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8500781655311584, "num_tokens": 173175810.0, "step": 4536 }, { "epoch": 0.5771530339651444, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.251118659973145, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8599545359611511, "num_tokens": 173220826.0, "step": 4537 }, { "epoch": 0.5772802442437349, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.257050514221191, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8700796961784363, "num_tokens": 173258411.0, "step": 4538 }, { "epoch": 0.5774074545223254, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.31460189819336, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8424766063690186, "num_tokens": 173297298.0, "step": 4539 }, { "epoch": 0.5775346648009159, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.259147644042969, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8421086668968201, "num_tokens": 173330839.0, "step": 4540 }, { "epoch": 0.5776618750795064, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.329405784606934, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8508371114730835, "num_tokens": 173370219.0, "step": 4541 }, { "epoch": 0.5777890853580969, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.272632598876953, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.858936071395874, "num_tokens": 173410425.0, "step": 4542 }, { "epoch": 0.5779162956366874, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.28222942352295, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8585638999938965, "num_tokens": 173447109.0, "step": 4543 }, { "epoch": 0.578043505915278, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.27676773071289, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8636504411697388, "num_tokens": 173490064.0, "step": 4544 }, { "epoch": 0.5781707161938685, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.318266868591309, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8608923554420471, "num_tokens": 173523791.0, "step": 4545 }, { "epoch": 0.5782979264724589, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.306907653808594, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8672361373901367, "num_tokens": 173564069.0, "step": 4546 }, { "epoch": 0.5784251367510495, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.310641288757324, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8579275608062744, "num_tokens": 173601893.0, "step": 4547 }, { "epoch": 0.57855234702964, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.275236129760742, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8683866262435913, "num_tokens": 173642299.0, "step": 4548 }, { "epoch": 0.5786795573082305, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.293048858642578, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8610007166862488, "num_tokens": 173675824.0, "step": 4549 }, { "epoch": 0.578806767586821, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.268031120300293, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8605895042419434, "num_tokens": 173712192.0, "step": 4550 }, { "epoch": 0.5789339778654116, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.302000999450684, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8663322329521179, "num_tokens": 173746949.0, "step": 4551 }, { "epoch": 0.579061188144002, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.283407211303711, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8615295886993408, "num_tokens": 173784527.0, "step": 4552 }, { "epoch": 0.5791883984225925, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.294678688049316, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8642934560775757, "num_tokens": 173820648.0, "step": 4553 }, { "epoch": 0.579315608701183, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.284727096557617, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8526872396469116, "num_tokens": 173860427.0, "step": 4554 }, { "epoch": 0.5794428189797736, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.261820793151855, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8610867261886597, "num_tokens": 173899876.0, "step": 4555 }, { "epoch": 0.5795700292583641, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.321098327636719, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8675412535667419, "num_tokens": 173937354.0, "step": 4556 }, { "epoch": 0.5796972395369546, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.30460262298584, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8473994731903076, "num_tokens": 173976039.0, "step": 4557 }, { "epoch": 0.5798244498155452, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.305739402770996, "learning_rate": 1e-06, "loss": 0.5076, "mean_token_accuracy": 0.8383768796920776, "num_tokens": 174006359.0, "step": 4558 }, { "epoch": 0.5799516600941356, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.33379077911377, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8510240912437439, "num_tokens": 174044662.0, "step": 4559 }, { "epoch": 0.5800788703727261, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.340591430664062, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8544266223907471, "num_tokens": 174077133.0, "step": 4560 }, { "epoch": 0.5802060806513166, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.255965232849121, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8542695045471191, "num_tokens": 174121873.0, "step": 4561 }, { "epoch": 0.5803332909299072, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.32545280456543, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.860334038734436, "num_tokens": 174158873.0, "step": 4562 }, { "epoch": 0.5804605012084977, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.312114715576172, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8655507564544678, "num_tokens": 174189222.0, "step": 4563 }, { "epoch": 0.5805877114870882, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.278932571411133, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8502459526062012, "num_tokens": 174229018.0, "step": 4564 }, { "epoch": 0.5807149217656786, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.30248737335205, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8596307039260864, "num_tokens": 174269215.0, "step": 4565 }, { "epoch": 0.5808421320442692, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.23733901977539, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.866976797580719, "num_tokens": 174303926.0, "step": 4566 }, { "epoch": 0.5809693423228597, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.253335952758789, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.861104428768158, "num_tokens": 174347491.0, "step": 4567 }, { "epoch": 0.5810965526014502, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.286422729492188, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.86103355884552, "num_tokens": 174384246.0, "step": 4568 }, { "epoch": 0.5812237628800407, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.282444953918457, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8573135137557983, "num_tokens": 174417081.0, "step": 4569 }, { "epoch": 0.5813509731586313, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.290876388549805, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8483593463897705, "num_tokens": 174457364.0, "step": 4570 }, { "epoch": 0.5814781834372217, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.278921127319336, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8599786758422852, "num_tokens": 174494304.0, "step": 4571 }, { "epoch": 0.5816053937158122, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.216253280639648, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8598626852035522, "num_tokens": 174528047.0, "step": 4572 }, { "epoch": 0.5817326039944027, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.282195091247559, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8641271591186523, "num_tokens": 174565990.0, "step": 4573 }, { "epoch": 0.5818598142729933, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.298948287963867, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8441460132598877, "num_tokens": 174599039.0, "step": 4574 }, { "epoch": 0.5819870245515838, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.27892780303955, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8524481654167175, "num_tokens": 174634786.0, "step": 4575 }, { "epoch": 0.5821142348301743, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.257193565368652, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8486654758453369, "num_tokens": 174677875.0, "step": 4576 }, { "epoch": 0.5822414451087647, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.293177604675293, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8525210618972778, "num_tokens": 174715421.0, "step": 4577 }, { "epoch": 0.5823686553873553, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.375662803649902, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8556350469589233, "num_tokens": 174746555.0, "step": 4578 }, { "epoch": 0.5824958656659458, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.267618179321289, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8475302457809448, "num_tokens": 174786470.0, "step": 4579 }, { "epoch": 0.5826230759445363, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.264967918395996, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8558626174926758, "num_tokens": 174827131.0, "step": 4580 }, { "epoch": 0.5827502862231269, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.346663475036621, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8516805768013, "num_tokens": 174861391.0, "step": 4581 }, { "epoch": 0.5828774965017174, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.258254051208496, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8625543117523193, "num_tokens": 174897437.0, "step": 4582 }, { "epoch": 0.5830047067803078, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.31855297088623, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8493866324424744, "num_tokens": 174932504.0, "step": 4583 }, { "epoch": 0.5831319170588983, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.303594589233398, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8525426387786865, "num_tokens": 174974746.0, "step": 4584 }, { "epoch": 0.5832591273374889, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.353556632995605, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8535752296447754, "num_tokens": 175016155.0, "step": 4585 }, { "epoch": 0.5833863376160794, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.281183242797852, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.842880368232727, "num_tokens": 175060595.0, "step": 4586 }, { "epoch": 0.5835135478946699, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.319893836975098, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8580033779144287, "num_tokens": 175101278.0, "step": 4587 }, { "epoch": 0.5836407581732604, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.301602363586426, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8512607216835022, "num_tokens": 175139479.0, "step": 4588 }, { "epoch": 0.5837679684518509, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.302053451538086, "learning_rate": 1e-06, "loss": 0.4908, "mean_token_accuracy": 0.8411524295806885, "num_tokens": 175182376.0, "step": 4589 }, { "epoch": 0.5838951787304414, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.412638664245605, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8505644202232361, "num_tokens": 175225449.0, "step": 4590 }, { "epoch": 0.5840223890090319, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.338114738464355, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8557550311088562, "num_tokens": 175255959.0, "step": 4591 }, { "epoch": 0.5841495992876224, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.289365768432617, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.845352053642273, "num_tokens": 175292142.0, "step": 4592 }, { "epoch": 0.584276809566213, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.275040626525879, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8516497611999512, "num_tokens": 175336630.0, "step": 4593 }, { "epoch": 0.5844040198448035, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.330011367797852, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8694829344749451, "num_tokens": 175372697.0, "step": 4594 }, { "epoch": 0.5845312301233939, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.287883758544922, "learning_rate": 1e-06, "loss": 0.4889, "mean_token_accuracy": 0.8451103568077087, "num_tokens": 175412395.0, "step": 4595 }, { "epoch": 0.5846584404019844, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.281367301940918, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8572636842727661, "num_tokens": 175447295.0, "step": 4596 }, { "epoch": 0.584785650680575, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.336256980895996, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8672293424606323, "num_tokens": 175480229.0, "step": 4597 }, { "epoch": 0.5849128609591655, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.274652481079102, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8574437499046326, "num_tokens": 175523081.0, "step": 4598 }, { "epoch": 0.585040071237756, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.292909622192383, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8611583709716797, "num_tokens": 175565036.0, "step": 4599 }, { "epoch": 0.5851672815163466, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.315104484558105, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8574374914169312, "num_tokens": 175606733.0, "step": 4600 }, { "epoch": 0.585294491794937, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.38255786895752, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8726189136505127, "num_tokens": 175637663.0, "step": 4601 }, { "epoch": 0.5854217020735275, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.319859504699707, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8601093292236328, "num_tokens": 175677381.0, "step": 4602 }, { "epoch": 0.585548912352118, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.316598892211914, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8689088821411133, "num_tokens": 175713049.0, "step": 4603 }, { "epoch": 0.5856761226307086, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.332326889038086, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8480185866355896, "num_tokens": 175753681.0, "step": 4604 }, { "epoch": 0.5858033329092991, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.306696891784668, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8455955386161804, "num_tokens": 175789725.0, "step": 4605 }, { "epoch": 0.5859305431878896, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.430489540100098, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8573172092437744, "num_tokens": 175823203.0, "step": 4606 }, { "epoch": 0.5860577534664801, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.287083625793457, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8525421619415283, "num_tokens": 175860506.0, "step": 4607 }, { "epoch": 0.5861849637450706, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.333529472351074, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.873437762260437, "num_tokens": 175895080.0, "step": 4608 }, { "epoch": 0.5863121740236611, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.323158264160156, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.851319432258606, "num_tokens": 175937113.0, "step": 4609 }, { "epoch": 0.5864393843022516, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.364415168762207, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.870509922504425, "num_tokens": 175966193.0, "step": 4610 }, { "epoch": 0.5865665945808421, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.321505546569824, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8515995144844055, "num_tokens": 176003771.0, "step": 4611 }, { "epoch": 0.5866938048594327, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.258540153503418, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8637669682502747, "num_tokens": 176050069.0, "step": 4612 }, { "epoch": 0.5868210151380232, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.297637939453125, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8723223209381104, "num_tokens": 176088763.0, "step": 4613 }, { "epoch": 0.5869482254166136, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.299976348876953, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.858771562576294, "num_tokens": 176127853.0, "step": 4614 }, { "epoch": 0.5870754356952042, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.296879768371582, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.843195915222168, "num_tokens": 176173775.0, "step": 4615 }, { "epoch": 0.5872026459737947, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.326532363891602, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8609185218811035, "num_tokens": 176210878.0, "step": 4616 }, { "epoch": 0.5873298562523852, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.284538269042969, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8522257804870605, "num_tokens": 176255094.0, "step": 4617 }, { "epoch": 0.5874570665309757, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.340940475463867, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8550764322280884, "num_tokens": 176293586.0, "step": 4618 }, { "epoch": 0.5875842768095663, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.320833206176758, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8600029945373535, "num_tokens": 176329203.0, "step": 4619 }, { "epoch": 0.5877114870881567, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.28563117980957, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.874824047088623, "num_tokens": 176367984.0, "step": 4620 }, { "epoch": 0.5878386973667472, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.321593284606934, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8716972470283508, "num_tokens": 176408949.0, "step": 4621 }, { "epoch": 0.5879659076453377, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.284994125366211, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8741875886917114, "num_tokens": 176451073.0, "step": 4622 }, { "epoch": 0.5880931179239283, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.311121940612793, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8656348586082458, "num_tokens": 176490623.0, "step": 4623 }, { "epoch": 0.5882203282025188, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.327048301696777, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8454856276512146, "num_tokens": 176525771.0, "step": 4624 }, { "epoch": 0.5883475384811093, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.255453109741211, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8762348890304565, "num_tokens": 176565417.0, "step": 4625 }, { "epoch": 0.5884747487596997, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.289022445678711, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8557679057121277, "num_tokens": 176599005.0, "step": 4626 }, { "epoch": 0.5886019590382903, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.323501586914062, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.861024022102356, "num_tokens": 176637947.0, "step": 4627 }, { "epoch": 0.5887291693168808, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.356806755065918, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8482843041419983, "num_tokens": 176675241.0, "step": 4628 }, { "epoch": 0.5888563795954713, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.311543464660645, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8515145778656006, "num_tokens": 176715375.0, "step": 4629 }, { "epoch": 0.5889835898740619, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.318503379821777, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8577596545219421, "num_tokens": 176750977.0, "step": 4630 }, { "epoch": 0.5891108001526524, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.274372100830078, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8621340990066528, "num_tokens": 176788968.0, "step": 4631 }, { "epoch": 0.5892380104312428, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.355818748474121, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8540443181991577, "num_tokens": 176824399.0, "step": 4632 }, { "epoch": 0.5893652207098333, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.305140495300293, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8539756536483765, "num_tokens": 176862053.0, "step": 4633 }, { "epoch": 0.5894924309884239, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.325509071350098, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8731346726417542, "num_tokens": 176896943.0, "step": 4634 }, { "epoch": 0.5896196412670144, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.267319679260254, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8591934442520142, "num_tokens": 176936263.0, "step": 4635 }, { "epoch": 0.5897468515456049, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.300898551940918, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8466090559959412, "num_tokens": 176976632.0, "step": 4636 }, { "epoch": 0.5898740618241954, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.358112335205078, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.864422082901001, "num_tokens": 177007036.0, "step": 4637 }, { "epoch": 0.5900012721027859, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.334822654724121, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8497461676597595, "num_tokens": 177042189.0, "step": 4638 }, { "epoch": 0.5901284823813764, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.335824012756348, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8539286851882935, "num_tokens": 177082428.0, "step": 4639 }, { "epoch": 0.5902556926599669, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.28125286102295, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8621296882629395, "num_tokens": 177124447.0, "step": 4640 }, { "epoch": 0.5903829029385574, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.325157165527344, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8743924498558044, "num_tokens": 177160516.0, "step": 4641 }, { "epoch": 0.590510113217148, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.282011985778809, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8672993183135986, "num_tokens": 177198100.0, "step": 4642 }, { "epoch": 0.5906373234957385, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.308747291564941, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8525235652923584, "num_tokens": 177243170.0, "step": 4643 }, { "epoch": 0.5907645337743289, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.297811508178711, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8565372824668884, "num_tokens": 177282976.0, "step": 4644 }, { "epoch": 0.5908917440529194, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.312623977661133, "learning_rate": 1e-06, "loss": 0.486, "mean_token_accuracy": 0.8415386080741882, "num_tokens": 177323725.0, "step": 4645 }, { "epoch": 0.59101895433151, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.306943893432617, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.858343780040741, "num_tokens": 177352178.0, "step": 4646 }, { "epoch": 0.5911461646101005, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.29791259765625, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8561147451400757, "num_tokens": 177392322.0, "step": 4647 }, { "epoch": 0.591273374888691, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.342164993286133, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8566348552703857, "num_tokens": 177432013.0, "step": 4648 }, { "epoch": 0.5914005851672816, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.29378604888916, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8723459243774414, "num_tokens": 177473542.0, "step": 4649 }, { "epoch": 0.591527795445872, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.285223960876465, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.855098307132721, "num_tokens": 177516590.0, "step": 4650 }, { "epoch": 0.5916550057244625, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.313156127929688, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8713473081588745, "num_tokens": 177553750.0, "step": 4651 }, { "epoch": 0.591782216003053, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.396429061889648, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8515430092811584, "num_tokens": 177586093.0, "step": 4652 }, { "epoch": 0.5919094262816436, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.28992748260498, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8548390865325928, "num_tokens": 177626800.0, "step": 4653 }, { "epoch": 0.5920366365602341, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.316969871520996, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8565775156021118, "num_tokens": 177662336.0, "step": 4654 }, { "epoch": 0.5921638468388246, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.281068801879883, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8662139177322388, "num_tokens": 177706922.0, "step": 4655 }, { "epoch": 0.592291057117415, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.359597206115723, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.856502890586853, "num_tokens": 177743922.0, "step": 4656 }, { "epoch": 0.5924182673960056, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.246711730957031, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8644980192184448, "num_tokens": 177777069.0, "step": 4657 }, { "epoch": 0.5925454776745961, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.357172012329102, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.867016077041626, "num_tokens": 177810959.0, "step": 4658 }, { "epoch": 0.5926726879531866, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.290445327758789, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8484994769096375, "num_tokens": 177846366.0, "step": 4659 }, { "epoch": 0.5927998982317771, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.37303352355957, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8709783554077148, "num_tokens": 177881450.0, "step": 4660 }, { "epoch": 0.5929271085103677, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.265307426452637, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8597303032875061, "num_tokens": 177922931.0, "step": 4661 }, { "epoch": 0.5930543187889582, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.303484916687012, "learning_rate": 1e-06, "loss": 0.4857, "mean_token_accuracy": 0.8405822515487671, "num_tokens": 177962722.0, "step": 4662 }, { "epoch": 0.5931815290675486, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.332770347595215, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.856711745262146, "num_tokens": 177999956.0, "step": 4663 }, { "epoch": 0.5933087393461391, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.312514305114746, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8476388454437256, "num_tokens": 178032450.0, "step": 4664 }, { "epoch": 0.5934359496247297, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.299711227416992, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.858948826789856, "num_tokens": 178075028.0, "step": 4665 }, { "epoch": 0.5935631599033202, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.33584976196289, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8547929525375366, "num_tokens": 178108267.0, "step": 4666 }, { "epoch": 0.5936903701819107, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.309025764465332, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8588663339614868, "num_tokens": 178148091.0, "step": 4667 }, { "epoch": 0.5938175804605013, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.302328109741211, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8714762926101685, "num_tokens": 178185626.0, "step": 4668 }, { "epoch": 0.5939447907390917, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.397476196289062, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8660140037536621, "num_tokens": 178224165.0, "step": 4669 }, { "epoch": 0.5940720010176822, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.27055549621582, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8576512336730957, "num_tokens": 178264118.0, "step": 4670 }, { "epoch": 0.5941992112962727, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.397313117980957, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8706152439117432, "num_tokens": 178295634.0, "step": 4671 }, { "epoch": 0.5943264215748633, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.375292778015137, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8649555444717407, "num_tokens": 178327076.0, "step": 4672 }, { "epoch": 0.5944536318534538, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.295926094055176, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8501993417739868, "num_tokens": 178357469.0, "step": 4673 }, { "epoch": 0.5945808421320443, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.386263847351074, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8647487163543701, "num_tokens": 178393830.0, "step": 4674 }, { "epoch": 0.5947080524106347, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.367368698120117, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8580745458602905, "num_tokens": 178427560.0, "step": 4675 }, { "epoch": 0.5948352626892253, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.31116008758545, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8601456880569458, "num_tokens": 178464666.0, "step": 4676 }, { "epoch": 0.5949624729678158, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.333152770996094, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8490692973136902, "num_tokens": 178508771.0, "step": 4677 }, { "epoch": 0.5950896832464063, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.313533782958984, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8626781702041626, "num_tokens": 178548516.0, "step": 4678 }, { "epoch": 0.5952168935249968, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.36562442779541, "learning_rate": 1e-06, "loss": 0.5559, "mean_token_accuracy": 0.8240326046943665, "num_tokens": 178591276.0, "step": 4679 }, { "epoch": 0.5953441038035874, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.465707778930664, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8578871488571167, "num_tokens": 178632580.0, "step": 4680 }, { "epoch": 0.5954713140821778, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.315717697143555, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8537729978561401, "num_tokens": 178672858.0, "step": 4681 }, { "epoch": 0.5955985243607683, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.387124061584473, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8593870401382446, "num_tokens": 178706435.0, "step": 4682 }, { "epoch": 0.5957257346393589, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.356975555419922, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8560285568237305, "num_tokens": 178746309.0, "step": 4683 }, { "epoch": 0.5958529449179494, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.28996467590332, "learning_rate": 1e-06, "loss": 0.4897, "mean_token_accuracy": 0.8409042358398438, "num_tokens": 178787639.0, "step": 4684 }, { "epoch": 0.5959801551965399, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.344268798828125, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8637988567352295, "num_tokens": 178828271.0, "step": 4685 }, { "epoch": 0.5961073654751304, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.349206924438477, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8794294595718384, "num_tokens": 178862862.0, "step": 4686 }, { "epoch": 0.5962345757537209, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.306800842285156, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8702536821365356, "num_tokens": 178897969.0, "step": 4687 }, { "epoch": 0.5963617860323114, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.274553298950195, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.863499104976654, "num_tokens": 178942002.0, "step": 4688 }, { "epoch": 0.5964889963109019, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.37634563446045, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8675543665885925, "num_tokens": 178983473.0, "step": 4689 }, { "epoch": 0.5966162065894924, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.299705505371094, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8473883867263794, "num_tokens": 179022370.0, "step": 4690 }, { "epoch": 0.596743416868083, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.275228500366211, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8740574717521667, "num_tokens": 179064841.0, "step": 4691 }, { "epoch": 0.5968706271466735, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.30543041229248, "learning_rate": 1e-06, "loss": 0.5169, "mean_token_accuracy": 0.8306299448013306, "num_tokens": 179099399.0, "step": 4692 }, { "epoch": 0.5969978374252639, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.3362398147583, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8497262001037598, "num_tokens": 179133969.0, "step": 4693 }, { "epoch": 0.5971250477038544, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.309342384338379, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8567346334457397, "num_tokens": 179165456.0, "step": 4694 }, { "epoch": 0.597252257982445, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.296189308166504, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8635613918304443, "num_tokens": 179198209.0, "step": 4695 }, { "epoch": 0.5973794682610355, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.291778564453125, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.843722939491272, "num_tokens": 179233872.0, "step": 4696 }, { "epoch": 0.597506678539626, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.269742012023926, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.8795802593231201, "num_tokens": 179272721.0, "step": 4697 }, { "epoch": 0.5976338888182166, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.29216480255127, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8567416667938232, "num_tokens": 179305781.0, "step": 4698 }, { "epoch": 0.597761099096807, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.30937671661377, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8630619049072266, "num_tokens": 179350049.0, "step": 4699 }, { "epoch": 0.5978883093753975, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.263128280639648, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8587247133255005, "num_tokens": 179388063.0, "step": 4700 }, { "epoch": 0.598015519653988, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.342903137207031, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8650697469711304, "num_tokens": 179427607.0, "step": 4701 }, { "epoch": 0.5981427299325786, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.309167861938477, "learning_rate": 1e-06, "loss": 0.4957, "mean_token_accuracy": 0.8415230512619019, "num_tokens": 179471456.0, "step": 4702 }, { "epoch": 0.5982699402111691, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.24215316772461, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8701473474502563, "num_tokens": 179514318.0, "step": 4703 }, { "epoch": 0.5983971504897596, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.398658752441406, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8713629245758057, "num_tokens": 179553307.0, "step": 4704 }, { "epoch": 0.59852436076835, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.305251121520996, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8508451581001282, "num_tokens": 179597288.0, "step": 4705 }, { "epoch": 0.5986515710469406, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.416768074035645, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8586430549621582, "num_tokens": 179629583.0, "step": 4706 }, { "epoch": 0.5987787813255311, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.333170890808105, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8532328605651855, "num_tokens": 179672104.0, "step": 4707 }, { "epoch": 0.5989059916041216, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.297518730163574, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8499369621276855, "num_tokens": 179712951.0, "step": 4708 }, { "epoch": 0.5990332018827121, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.325435638427734, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8578978776931763, "num_tokens": 179754626.0, "step": 4709 }, { "epoch": 0.5991604121613027, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.308303833007812, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8641715049743652, "num_tokens": 179790068.0, "step": 4710 }, { "epoch": 0.5992876224398932, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.369230270385742, "learning_rate": 1e-06, "loss": 0.4995, "mean_token_accuracy": 0.8393270969390869, "num_tokens": 179823936.0, "step": 4711 }, { "epoch": 0.5994148327184836, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.31850528717041, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8601775169372559, "num_tokens": 179860920.0, "step": 4712 }, { "epoch": 0.5995420429970741, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.354120254516602, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8495872020721436, "num_tokens": 179899848.0, "step": 4713 }, { "epoch": 0.5996692532756647, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.33203125, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8750531673431396, "num_tokens": 179934462.0, "step": 4714 }, { "epoch": 0.5997964635542552, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.291300773620605, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8592761158943176, "num_tokens": 179973955.0, "step": 4715 }, { "epoch": 0.5999236738328457, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.34104061126709, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8604583144187927, "num_tokens": 180015256.0, "step": 4716 }, { "epoch": 0.6000508841114363, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.279940605163574, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8665560483932495, "num_tokens": 180054145.0, "step": 4717 }, { "epoch": 0.6001780943900267, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.316793441772461, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8633831143379211, "num_tokens": 180093411.0, "step": 4718 }, { "epoch": 0.6003053046686172, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.291150093078613, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8682692646980286, "num_tokens": 180135799.0, "step": 4719 }, { "epoch": 0.6004325149472077, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.3167724609375, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.853488564491272, "num_tokens": 180172623.0, "step": 4720 }, { "epoch": 0.6005597252257983, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.364436149597168, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8552093505859375, "num_tokens": 180213833.0, "step": 4721 }, { "epoch": 0.6006869355043888, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.285680770874023, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8664271831512451, "num_tokens": 180254982.0, "step": 4722 }, { "epoch": 0.6008141457829793, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.335052490234375, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8673632144927979, "num_tokens": 180288544.0, "step": 4723 }, { "epoch": 0.6009413560615697, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.366320610046387, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8557949066162109, "num_tokens": 180330233.0, "step": 4724 }, { "epoch": 0.6010685663401603, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.326781272888184, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8615397214889526, "num_tokens": 180361516.0, "step": 4725 }, { "epoch": 0.6011957766187508, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.347911834716797, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8574985861778259, "num_tokens": 180404380.0, "step": 4726 }, { "epoch": 0.6013229868973413, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.373763084411621, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8524646759033203, "num_tokens": 180436916.0, "step": 4727 }, { "epoch": 0.6014501971759318, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.318831443786621, "learning_rate": 1e-06, "loss": 0.4926, "mean_token_accuracy": 0.841417670249939, "num_tokens": 180476895.0, "step": 4728 }, { "epoch": 0.6015774074545224, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.334663391113281, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8592247366905212, "num_tokens": 180515248.0, "step": 4729 }, { "epoch": 0.6017046177331128, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.330751419067383, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8654007911682129, "num_tokens": 180554575.0, "step": 4730 }, { "epoch": 0.6018318280117033, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.36768627166748, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8493664860725403, "num_tokens": 180594485.0, "step": 4731 }, { "epoch": 0.6019590382902938, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.344194412231445, "learning_rate": 1e-06, "loss": 0.4808, "mean_token_accuracy": 0.8442538976669312, "num_tokens": 180634746.0, "step": 4732 }, { "epoch": 0.6020862485688844, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.379484176635742, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.849866509437561, "num_tokens": 180674946.0, "step": 4733 }, { "epoch": 0.6022134588474749, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.339295387268066, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8585944771766663, "num_tokens": 180715538.0, "step": 4734 }, { "epoch": 0.6023406691260654, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.309090614318848, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8574408292770386, "num_tokens": 180755298.0, "step": 4735 }, { "epoch": 0.6024678794046558, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.32138729095459, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8531985282897949, "num_tokens": 180797104.0, "step": 4736 }, { "epoch": 0.6025950896832464, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.336319923400879, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8693825006484985, "num_tokens": 180839845.0, "step": 4737 }, { "epoch": 0.6027222999618369, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.313887596130371, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8735851645469666, "num_tokens": 180878849.0, "step": 4738 }, { "epoch": 0.6028495102404274, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.411711692810059, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8578532934188843, "num_tokens": 180918427.0, "step": 4739 }, { "epoch": 0.602976720519018, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.307320594787598, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8517624139785767, "num_tokens": 180956206.0, "step": 4740 }, { "epoch": 0.6031039307976085, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.350249290466309, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8646743297576904, "num_tokens": 180996945.0, "step": 4741 }, { "epoch": 0.6032311410761989, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.355202674865723, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8644900321960449, "num_tokens": 181031763.0, "step": 4742 }, { "epoch": 0.6033583513547894, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.281831741333008, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8668838143348694, "num_tokens": 181070661.0, "step": 4743 }, { "epoch": 0.60348556163338, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.281294822692871, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8453909754753113, "num_tokens": 181112769.0, "step": 4744 }, { "epoch": 0.6036127719119705, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.318110466003418, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8591313362121582, "num_tokens": 181151155.0, "step": 4745 }, { "epoch": 0.603739982190561, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.32997989654541, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8588491082191467, "num_tokens": 181188689.0, "step": 4746 }, { "epoch": 0.6038671924691515, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.319550514221191, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8633556365966797, "num_tokens": 181230031.0, "step": 4747 }, { "epoch": 0.603994402747742, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.328672409057617, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8595119714736938, "num_tokens": 181271464.0, "step": 4748 }, { "epoch": 0.6041216130263325, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.324594497680664, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8536199331283569, "num_tokens": 181312607.0, "step": 4749 }, { "epoch": 0.604248823304923, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.338658332824707, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8675199747085571, "num_tokens": 181348628.0, "step": 4750 }, { "epoch": 0.6043760335835135, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.313398361206055, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8512008786201477, "num_tokens": 181387975.0, "step": 4751 }, { "epoch": 0.6045032438621041, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.34221363067627, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8717517256736755, "num_tokens": 181424163.0, "step": 4752 }, { "epoch": 0.6046304541406946, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.333047866821289, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8492714166641235, "num_tokens": 181457002.0, "step": 4753 }, { "epoch": 0.604757664419285, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.399086952209473, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8545360565185547, "num_tokens": 181494410.0, "step": 4754 }, { "epoch": 0.6048848746978756, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.285418510437012, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8564345240592957, "num_tokens": 181534595.0, "step": 4755 }, { "epoch": 0.6050120849764661, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.374310493469238, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.865452766418457, "num_tokens": 181570654.0, "step": 4756 }, { "epoch": 0.6051392952550566, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.323668479919434, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8539307117462158, "num_tokens": 181608366.0, "step": 4757 }, { "epoch": 0.6052665055336471, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.359589576721191, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8471658229827881, "num_tokens": 181647168.0, "step": 4758 }, { "epoch": 0.6053937158122377, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.360372543334961, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8624498844146729, "num_tokens": 181689177.0, "step": 4759 }, { "epoch": 0.6055209260908282, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.352214813232422, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8639689683914185, "num_tokens": 181723742.0, "step": 4760 }, { "epoch": 0.6056481363694186, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.349099159240723, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.867368221282959, "num_tokens": 181761539.0, "step": 4761 }, { "epoch": 0.6057753466480091, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.405091285705566, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8567301034927368, "num_tokens": 181799674.0, "step": 4762 }, { "epoch": 0.6059025569265997, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.358048439025879, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8469394445419312, "num_tokens": 181838784.0, "step": 4763 }, { "epoch": 0.6060297672051902, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.374863624572754, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8546115159988403, "num_tokens": 181875669.0, "step": 4764 }, { "epoch": 0.6061569774837807, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.34011173248291, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8422024250030518, "num_tokens": 181912450.0, "step": 4765 }, { "epoch": 0.6062841877623713, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.35947036743164, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8589016199111938, "num_tokens": 181947691.0, "step": 4766 }, { "epoch": 0.6064113980409617, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.427879333496094, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8636221885681152, "num_tokens": 181982364.0, "step": 4767 }, { "epoch": 0.6065386083195522, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.398138046264648, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8587148189544678, "num_tokens": 182017071.0, "step": 4768 }, { "epoch": 0.6066658185981427, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.34105110168457, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8587490320205688, "num_tokens": 182057234.0, "step": 4769 }, { "epoch": 0.6067930288767333, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.324565887451172, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8696756362915039, "num_tokens": 182092450.0, "step": 4770 }, { "epoch": 0.6069202391553238, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.42174243927002, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8507599830627441, "num_tokens": 182126398.0, "step": 4771 }, { "epoch": 0.6070474494339143, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.324832916259766, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8739388585090637, "num_tokens": 182160848.0, "step": 4772 }, { "epoch": 0.6071746597125047, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.293213844299316, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8674182891845703, "num_tokens": 182202596.0, "step": 4773 }, { "epoch": 0.6073018699910953, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.365331649780273, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8577207326889038, "num_tokens": 182239235.0, "step": 4774 }, { "epoch": 0.6074290802696858, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.376005172729492, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8713377714157104, "num_tokens": 182273662.0, "step": 4775 }, { "epoch": 0.6075562905482763, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.396208763122559, "learning_rate": 1e-06, "loss": 0.4936, "mean_token_accuracy": 0.8446370363235474, "num_tokens": 182311991.0, "step": 4776 }, { "epoch": 0.6076835008268668, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.373955726623535, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8610196113586426, "num_tokens": 182350722.0, "step": 4777 }, { "epoch": 0.6078107111054574, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.368949890136719, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.861474871635437, "num_tokens": 182386755.0, "step": 4778 }, { "epoch": 0.6079379213840478, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.4179105758667, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8753442168235779, "num_tokens": 182423128.0, "step": 4779 }, { "epoch": 0.6080651316626383, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.337658882141113, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8456568717956543, "num_tokens": 182461518.0, "step": 4780 }, { "epoch": 0.6081923419412288, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.396951675415039, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8595302104949951, "num_tokens": 182495707.0, "step": 4781 }, { "epoch": 0.6083195522198194, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.350836753845215, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.849927544593811, "num_tokens": 182539058.0, "step": 4782 }, { "epoch": 0.6084467624984099, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.45889949798584, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8455633521080017, "num_tokens": 182572482.0, "step": 4783 }, { "epoch": 0.6085739727770004, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.394433975219727, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8632642030715942, "num_tokens": 182607665.0, "step": 4784 }, { "epoch": 0.6087011830555908, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.371387481689453, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8607370853424072, "num_tokens": 182650503.0, "step": 4785 }, { "epoch": 0.6088283933341814, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.420381546020508, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8699032664299011, "num_tokens": 182681662.0, "step": 4786 }, { "epoch": 0.6089556036127719, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.391592025756836, "learning_rate": 1e-06, "loss": 0.4783, "mean_token_accuracy": 0.8462189435958862, "num_tokens": 182714912.0, "step": 4787 }, { "epoch": 0.6090828138913624, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.378914833068848, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8569535613059998, "num_tokens": 182748813.0, "step": 4788 }, { "epoch": 0.609210024169953, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.357704162597656, "learning_rate": 1e-06, "loss": 0.4877, "mean_token_accuracy": 0.8436645269393921, "num_tokens": 182786123.0, "step": 4789 }, { "epoch": 0.6093372344485435, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.3790922164917, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8725574016571045, "num_tokens": 182820550.0, "step": 4790 }, { "epoch": 0.6094644447271339, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.394820213317871, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8604341149330139, "num_tokens": 182863135.0, "step": 4791 }, { "epoch": 0.6095916550057244, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.340179443359375, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8612809181213379, "num_tokens": 182905193.0, "step": 4792 }, { "epoch": 0.609718865284315, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.322152137756348, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8586739301681519, "num_tokens": 182944934.0, "step": 4793 }, { "epoch": 0.6098460755629055, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.33924388885498, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.8462307453155518, "num_tokens": 182984604.0, "step": 4794 }, { "epoch": 0.609973285841496, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.398529052734375, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8578604459762573, "num_tokens": 183020421.0, "step": 4795 }, { "epoch": 0.6101004961200865, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.353211402893066, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8514584302902222, "num_tokens": 183057635.0, "step": 4796 }, { "epoch": 0.610227706398677, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.32135009765625, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8469133377075195, "num_tokens": 183094689.0, "step": 4797 }, { "epoch": 0.6103549166772675, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.414501190185547, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8645573258399963, "num_tokens": 183129118.0, "step": 4798 }, { "epoch": 0.610482126955858, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.344644546508789, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8574583530426025, "num_tokens": 183163365.0, "step": 4799 }, { "epoch": 0.6106093372344485, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.35130786895752, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8633044958114624, "num_tokens": 183199271.0, "step": 4800 }, { "epoch": 0.6107365475130391, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.381417274475098, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8787358403205872, "num_tokens": 183234523.0, "step": 4801 }, { "epoch": 0.6108637577916296, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.371419906616211, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8592524528503418, "num_tokens": 183267098.0, "step": 4802 }, { "epoch": 0.61099096807022, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.406922340393066, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8596221208572388, "num_tokens": 183306825.0, "step": 4803 }, { "epoch": 0.6111181783488105, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.361623764038086, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8587212562561035, "num_tokens": 183348712.0, "step": 4804 }, { "epoch": 0.6112453886274011, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.40966796875, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8707882761955261, "num_tokens": 183383797.0, "step": 4805 }, { "epoch": 0.6113725989059916, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.460890769958496, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8705430030822754, "num_tokens": 183417484.0, "step": 4806 }, { "epoch": 0.6114998091845821, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.415069580078125, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8654743432998657, "num_tokens": 183453808.0, "step": 4807 }, { "epoch": 0.6116270194631727, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.496467590332031, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8665449023246765, "num_tokens": 183496693.0, "step": 4808 }, { "epoch": 0.6117542297417632, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.329313278198242, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.862229585647583, "num_tokens": 183536727.0, "step": 4809 }, { "epoch": 0.6118814400203536, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.497868537902832, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.861923336982727, "num_tokens": 183571809.0, "step": 4810 }, { "epoch": 0.6120086502989441, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.426714897155762, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8586289286613464, "num_tokens": 183610761.0, "step": 4811 }, { "epoch": 0.6121358605775347, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.393856048583984, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8581714630126953, "num_tokens": 183648304.0, "step": 4812 }, { "epoch": 0.6122630708561252, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.444938659667969, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.855673611164093, "num_tokens": 183684201.0, "step": 4813 }, { "epoch": 0.6123902811347157, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.409372329711914, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8525350689888, "num_tokens": 183726595.0, "step": 4814 }, { "epoch": 0.6125174914133062, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.406946182250977, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8683146238327026, "num_tokens": 183762222.0, "step": 4815 }, { "epoch": 0.6126447016918967, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.443367004394531, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8722782135009766, "num_tokens": 183797167.0, "step": 4816 }, { "epoch": 0.6127719119704872, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.37822437286377, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8578125238418579, "num_tokens": 183839591.0, "step": 4817 }, { "epoch": 0.6128991222490777, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.310763359069824, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8599108457565308, "num_tokens": 183877862.0, "step": 4818 }, { "epoch": 0.6130263325276682, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.375229835510254, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8652568459510803, "num_tokens": 183915964.0, "step": 4819 }, { "epoch": 0.6131535428062588, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.438976287841797, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8563219904899597, "num_tokens": 183952823.0, "step": 4820 }, { "epoch": 0.6132807530848493, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.368599891662598, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.850403904914856, "num_tokens": 183991650.0, "step": 4821 }, { "epoch": 0.6134079633634397, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.407362937927246, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8611979484558105, "num_tokens": 184027768.0, "step": 4822 }, { "epoch": 0.6135351736420303, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.419878959655762, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8719479441642761, "num_tokens": 184063754.0, "step": 4823 }, { "epoch": 0.6136623839206208, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.389298439025879, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8539698123931885, "num_tokens": 184105286.0, "step": 4824 }, { "epoch": 0.6137895941992113, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.361099243164062, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8648160099983215, "num_tokens": 184139312.0, "step": 4825 }, { "epoch": 0.6139168044778018, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.351685523986816, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.858708381652832, "num_tokens": 184181541.0, "step": 4826 }, { "epoch": 0.6140440147563924, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.365574836730957, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8593829870223999, "num_tokens": 184216857.0, "step": 4827 }, { "epoch": 0.6141712250349828, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.364120483398438, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8635278940200806, "num_tokens": 184253318.0, "step": 4828 }, { "epoch": 0.6142984353135733, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.387887001037598, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8711341619491577, "num_tokens": 184298625.0, "step": 4829 }, { "epoch": 0.6144256455921638, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.3956937789917, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8578738570213318, "num_tokens": 184334569.0, "step": 4830 }, { "epoch": 0.6145528558707544, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.343220710754395, "learning_rate": 1e-06, "loss": 0.4949, "mean_token_accuracy": 0.8427672386169434, "num_tokens": 184377477.0, "step": 4831 }, { "epoch": 0.6146800661493449, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.370972633361816, "learning_rate": 1e-06, "loss": 0.4914, "mean_token_accuracy": 0.8464189171791077, "num_tokens": 184416671.0, "step": 4832 }, { "epoch": 0.6148072764279354, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.39454460144043, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8546299934387207, "num_tokens": 184453174.0, "step": 4833 }, { "epoch": 0.6149344867065258, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.383187294006348, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8661817908287048, "num_tokens": 184488313.0, "step": 4834 }, { "epoch": 0.6150616969851164, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.452954292297363, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8400633335113525, "num_tokens": 184523733.0, "step": 4835 }, { "epoch": 0.6151889072637069, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.34512996673584, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8569574952125549, "num_tokens": 184568399.0, "step": 4836 }, { "epoch": 0.6153161175422974, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.45447826385498, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.858422040939331, "num_tokens": 184603302.0, "step": 4837 }, { "epoch": 0.615443327820888, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.437993049621582, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8793313503265381, "num_tokens": 184637346.0, "step": 4838 }, { "epoch": 0.6155705380994785, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.433865547180176, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.866321325302124, "num_tokens": 184671687.0, "step": 4839 }, { "epoch": 0.6156977483780689, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.409218788146973, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8689911365509033, "num_tokens": 184706867.0, "step": 4840 }, { "epoch": 0.6158249586566594, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.46357536315918, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8557131886482239, "num_tokens": 184743916.0, "step": 4841 }, { "epoch": 0.61595216893525, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.29405689239502, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8687247037887573, "num_tokens": 184783247.0, "step": 4842 }, { "epoch": 0.6160793792138405, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.444832801818848, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.859707772731781, "num_tokens": 184826501.0, "step": 4843 }, { "epoch": 0.616206589492431, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.393261909484863, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8578811883926392, "num_tokens": 184868743.0, "step": 4844 }, { "epoch": 0.6163337997710215, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.37448787689209, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8558574914932251, "num_tokens": 184908992.0, "step": 4845 }, { "epoch": 0.616461010049612, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.485048294067383, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8481093645095825, "num_tokens": 184942456.0, "step": 4846 }, { "epoch": 0.6165882203282025, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.420151710510254, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8505620956420898, "num_tokens": 184975834.0, "step": 4847 }, { "epoch": 0.616715430606793, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.368227005004883, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.881088137626648, "num_tokens": 185012295.0, "step": 4848 }, { "epoch": 0.6168426408853835, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.478251457214355, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8643462657928467, "num_tokens": 185049272.0, "step": 4849 }, { "epoch": 0.6169698511639741, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.386296272277832, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8718613982200623, "num_tokens": 185081054.0, "step": 4850 }, { "epoch": 0.6170970614425646, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.431557655334473, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8555850386619568, "num_tokens": 185119809.0, "step": 4851 }, { "epoch": 0.617224271721155, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.33759593963623, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8711700439453125, "num_tokens": 185161462.0, "step": 4852 }, { "epoch": 0.6173514819997455, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.428316116333008, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8510198593139648, "num_tokens": 185196238.0, "step": 4853 }, { "epoch": 0.6174786922783361, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.41338062286377, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.859169602394104, "num_tokens": 185239317.0, "step": 4854 }, { "epoch": 0.6176059025569266, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.361544609069824, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8689455986022949, "num_tokens": 185273555.0, "step": 4855 }, { "epoch": 0.6177331128355171, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.395419120788574, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8674772381782532, "num_tokens": 185309721.0, "step": 4856 }, { "epoch": 0.6178603231141077, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.393366813659668, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8527234196662903, "num_tokens": 185350846.0, "step": 4857 }, { "epoch": 0.6179875333926982, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.398173332214355, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8523027896881104, "num_tokens": 185392472.0, "step": 4858 }, { "epoch": 0.6181147436712886, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.425588607788086, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8558782339096069, "num_tokens": 185438250.0, "step": 4859 }, { "epoch": 0.6182419539498791, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.369282722473145, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.856525719165802, "num_tokens": 185479455.0, "step": 4860 }, { "epoch": 0.6183691642284697, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.457503318786621, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8561277985572815, "num_tokens": 185518719.0, "step": 4861 }, { "epoch": 0.6184963745070602, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.382367134094238, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.845594584941864, "num_tokens": 185556671.0, "step": 4862 }, { "epoch": 0.6186235847856507, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.382413864135742, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8566259145736694, "num_tokens": 185595063.0, "step": 4863 }, { "epoch": 0.6187507950642412, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.415759086608887, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8495908379554749, "num_tokens": 185635626.0, "step": 4864 }, { "epoch": 0.6188780053428317, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.476123809814453, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8477439880371094, "num_tokens": 185671687.0, "step": 4865 }, { "epoch": 0.6190052156214222, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.36608600616455, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.863100528717041, "num_tokens": 185706459.0, "step": 4866 }, { "epoch": 0.6191324259000127, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.428503036499023, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8778802752494812, "num_tokens": 185743162.0, "step": 4867 }, { "epoch": 0.6192596361786032, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.435521125793457, "learning_rate": 1e-06, "loss": 0.4892, "mean_token_accuracy": 0.844115138053894, "num_tokens": 185784036.0, "step": 4868 }, { "epoch": 0.6193868464571938, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.38002872467041, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8564451932907104, "num_tokens": 185821407.0, "step": 4869 }, { "epoch": 0.6195140567357843, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.456733703613281, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8647168874740601, "num_tokens": 185857789.0, "step": 4870 }, { "epoch": 0.6196412670143747, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.429339408874512, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.870116114616394, "num_tokens": 185896989.0, "step": 4871 }, { "epoch": 0.6197684772929652, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.372902870178223, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8738176822662354, "num_tokens": 185938123.0, "step": 4872 }, { "epoch": 0.6198956875715558, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.446748733520508, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8603850603103638, "num_tokens": 185973440.0, "step": 4873 }, { "epoch": 0.6200228978501463, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.353739738464355, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8597387075424194, "num_tokens": 186010069.0, "step": 4874 }, { "epoch": 0.6201501081287368, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.449520111083984, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8542011976242065, "num_tokens": 186048365.0, "step": 4875 }, { "epoch": 0.6202773184073274, "ewc_loss": 0.018798828125, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.39541244506836, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8646819591522217, "num_tokens": 186085358.0, "step": 4876 }, { "epoch": 0.6204045286859178, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.433114051818848, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8635273575782776, "num_tokens": 186118476.0, "step": 4877 }, { "epoch": 0.6205317389645083, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.425469398498535, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8555341958999634, "num_tokens": 186155487.0, "step": 4878 }, { "epoch": 0.6206589492430988, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.41185188293457, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8653860688209534, "num_tokens": 186192941.0, "step": 4879 }, { "epoch": 0.6207861595216894, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.368595123291016, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8630123138427734, "num_tokens": 186233860.0, "step": 4880 }, { "epoch": 0.6209133698002799, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.444633483886719, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8483841419219971, "num_tokens": 186273081.0, "step": 4881 }, { "epoch": 0.6210405800788704, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.394599914550781, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.865568995475769, "num_tokens": 186310596.0, "step": 4882 }, { "epoch": 0.6211677903574608, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.468273162841797, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8556132316589355, "num_tokens": 186346262.0, "step": 4883 }, { "epoch": 0.6212950006360514, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.374608993530273, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8620059490203857, "num_tokens": 186386508.0, "step": 4884 }, { "epoch": 0.6214222109146419, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.377666473388672, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8554548621177673, "num_tokens": 186425184.0, "step": 4885 }, { "epoch": 0.6215494211932324, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.388643264770508, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8585390448570251, "num_tokens": 186463072.0, "step": 4886 }, { "epoch": 0.621676631471823, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.388256072998047, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8701193928718567, "num_tokens": 186504062.0, "step": 4887 }, { "epoch": 0.6218038417504135, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.418582916259766, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8499250411987305, "num_tokens": 186544564.0, "step": 4888 }, { "epoch": 0.6219310520290039, "ewc_loss": 0.018798828125, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.372888565063477, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8464210033416748, "num_tokens": 186587196.0, "step": 4889 }, { "epoch": 0.6220582623075944, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.40578556060791, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8600289821624756, "num_tokens": 186620673.0, "step": 4890 }, { "epoch": 0.622185472586185, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.378908157348633, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.861747145652771, "num_tokens": 186654028.0, "step": 4891 }, { "epoch": 0.6223126828647755, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.392471313476562, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8562102913856506, "num_tokens": 186693395.0, "step": 4892 }, { "epoch": 0.622439893143366, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.3712158203125, "learning_rate": 1e-06, "loss": 0.4934, "mean_token_accuracy": 0.8418856859207153, "num_tokens": 186734440.0, "step": 4893 }, { "epoch": 0.6225671034219565, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.363686561584473, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8622838854789734, "num_tokens": 186769637.0, "step": 4894 }, { "epoch": 0.622694313700547, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.433140754699707, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8553237318992615, "num_tokens": 186810188.0, "step": 4895 }, { "epoch": 0.6228215239791375, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.443330764770508, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8496497869491577, "num_tokens": 186847512.0, "step": 4896 }, { "epoch": 0.622948734257728, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.371932029724121, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8686729669570923, "num_tokens": 186878721.0, "step": 4897 }, { "epoch": 0.6230759445363185, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.4088134765625, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8643734455108643, "num_tokens": 186920239.0, "step": 4898 }, { "epoch": 0.6232031548149091, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.341050148010254, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8529706001281738, "num_tokens": 186964158.0, "step": 4899 }, { "epoch": 0.6233303650934996, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.38880443572998, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8740660548210144, "num_tokens": 187003176.0, "step": 4900 }, { "epoch": 0.62345757537209, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.32856559753418, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.858834981918335, "num_tokens": 187046152.0, "step": 4901 }, { "epoch": 0.6235847856506805, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.346147537231445, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8717901706695557, "num_tokens": 187086629.0, "step": 4902 }, { "epoch": 0.6237119959292711, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.400284767150879, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8538651466369629, "num_tokens": 187122856.0, "step": 4903 }, { "epoch": 0.6238392062078616, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.488531112670898, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8738536834716797, "num_tokens": 187161568.0, "step": 4904 }, { "epoch": 0.6239664164864521, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.396500587463379, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8605093955993652, "num_tokens": 187202831.0, "step": 4905 }, { "epoch": 0.6240936267650427, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.358116149902344, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8665924668312073, "num_tokens": 187239059.0, "step": 4906 }, { "epoch": 0.6242208370436332, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.450128555297852, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8537485003471375, "num_tokens": 187272940.0, "step": 4907 }, { "epoch": 0.6243480473222236, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.425164222717285, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8723267316818237, "num_tokens": 187310963.0, "step": 4908 }, { "epoch": 0.6244752576008141, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.3831148147583, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8785483837127686, "num_tokens": 187345698.0, "step": 4909 }, { "epoch": 0.6246024678794047, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.40854263305664, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8580406904220581, "num_tokens": 187384762.0, "step": 4910 }, { "epoch": 0.6247296781579952, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.422088623046875, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8493733406066895, "num_tokens": 187424288.0, "step": 4911 }, { "epoch": 0.6248568884365857, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.376280784606934, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8686626553535461, "num_tokens": 187461481.0, "step": 4912 }, { "epoch": 0.6249840987151762, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.490050315856934, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8747446537017822, "num_tokens": 187497690.0, "step": 4913 }, { "epoch": 0.6251113089937667, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.33692741394043, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8539823293685913, "num_tokens": 187537825.0, "step": 4914 }, { "epoch": 0.6252385192723572, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.3668794631958, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8701596856117249, "num_tokens": 187583998.0, "step": 4915 }, { "epoch": 0.6253657295509477, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.360784530639648, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8589516282081604, "num_tokens": 187621962.0, "step": 4916 }, { "epoch": 0.6254929398295382, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.464611053466797, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8668696880340576, "num_tokens": 187655913.0, "step": 4917 }, { "epoch": 0.6256201501081288, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.377592086791992, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8517231941223145, "num_tokens": 187692433.0, "step": 4918 }, { "epoch": 0.6257473603867193, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.4322509765625, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8721627593040466, "num_tokens": 187724311.0, "step": 4919 }, { "epoch": 0.6258745706653097, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.38473129272461, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8685003519058228, "num_tokens": 187766614.0, "step": 4920 }, { "epoch": 0.6260017809439002, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.381397247314453, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8708493709564209, "num_tokens": 187804847.0, "step": 4921 }, { "epoch": 0.6261289912224908, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.394079208374023, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8620038032531738, "num_tokens": 187842791.0, "step": 4922 }, { "epoch": 0.6262562015010813, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.395964622497559, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8504047989845276, "num_tokens": 187878304.0, "step": 4923 }, { "epoch": 0.6263834117796718, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.382133483886719, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8626224994659424, "num_tokens": 187915022.0, "step": 4924 }, { "epoch": 0.6265106220582624, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.401150703430176, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8609330058097839, "num_tokens": 187950533.0, "step": 4925 }, { "epoch": 0.6266378323368528, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.402087211608887, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8558906316757202, "num_tokens": 187989893.0, "step": 4926 }, { "epoch": 0.6267650426154433, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.40009880065918, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8528682589530945, "num_tokens": 188025031.0, "step": 4927 }, { "epoch": 0.6268922528940338, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.41549301147461, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.852716326713562, "num_tokens": 188060797.0, "step": 4928 }, { "epoch": 0.6270194631726244, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.450618743896484, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8528209328651428, "num_tokens": 188096271.0, "step": 4929 }, { "epoch": 0.6271466734512149, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.401768684387207, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8605192303657532, "num_tokens": 188136697.0, "step": 4930 }, { "epoch": 0.6272738837298054, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.38029670715332, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8578217029571533, "num_tokens": 188174715.0, "step": 4931 }, { "epoch": 0.6274010940083958, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.418529510498047, "learning_rate": 1e-06, "loss": 0.5156, "mean_token_accuracy": 0.8384174704551697, "num_tokens": 188216379.0, "step": 4932 }, { "epoch": 0.6275283042869864, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.384613990783691, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8649589419364929, "num_tokens": 188252183.0, "step": 4933 }, { "epoch": 0.6276555145655769, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.448091506958008, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8626218438148499, "num_tokens": 188291128.0, "step": 4934 }, { "epoch": 0.6277827248441674, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.3732271194458, "learning_rate": 1e-06, "loss": 0.5015, "mean_token_accuracy": 0.8429533243179321, "num_tokens": 188334256.0, "step": 4935 }, { "epoch": 0.627909935122758, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.415000915527344, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8523538708686829, "num_tokens": 188373207.0, "step": 4936 }, { "epoch": 0.6280371454013485, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.479552268981934, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8560330867767334, "num_tokens": 188414937.0, "step": 4937 }, { "epoch": 0.6281643556799389, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.37451457977295, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8561972379684448, "num_tokens": 188455398.0, "step": 4938 }, { "epoch": 0.6282915659585294, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.380900382995605, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8588801026344299, "num_tokens": 188496561.0, "step": 4939 }, { "epoch": 0.62841877623712, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.4556245803833, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8635262846946716, "num_tokens": 188531318.0, "step": 4940 }, { "epoch": 0.6285459865157105, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.369004249572754, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8509238958358765, "num_tokens": 188566807.0, "step": 4941 }, { "epoch": 0.628673196794301, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.405487060546875, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8559012413024902, "num_tokens": 188608062.0, "step": 4942 }, { "epoch": 0.6288004070728915, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.376663208007812, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8710166811943054, "num_tokens": 188644953.0, "step": 4943 }, { "epoch": 0.628927617351482, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.404537200927734, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8748199343681335, "num_tokens": 188687720.0, "step": 4944 }, { "epoch": 0.6290548276300725, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.436391830444336, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.852603018283844, "num_tokens": 188730077.0, "step": 4945 }, { "epoch": 0.629182037908663, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.477477073669434, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8660085201263428, "num_tokens": 188768306.0, "step": 4946 }, { "epoch": 0.6293092481872535, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.415102005004883, "learning_rate": 1e-06, "loss": 0.502, "mean_token_accuracy": 0.8389883041381836, "num_tokens": 188807480.0, "step": 4947 }, { "epoch": 0.6294364584658441, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.500216484069824, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.850213885307312, "num_tokens": 188836383.0, "step": 4948 }, { "epoch": 0.6295636687444346, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.464813232421875, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8627657890319824, "num_tokens": 188876079.0, "step": 4949 }, { "epoch": 0.629690879023025, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.408493995666504, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8569964170455933, "num_tokens": 188912936.0, "step": 4950 }, { "epoch": 0.6298180893016155, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.435962677001953, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8583855032920837, "num_tokens": 188951475.0, "step": 4951 }, { "epoch": 0.6299452995802061, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.385696411132812, "learning_rate": 1e-06, "loss": 0.4875, "mean_token_accuracy": 0.844948410987854, "num_tokens": 188996972.0, "step": 4952 }, { "epoch": 0.6300725098587966, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.437960624694824, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8788855075836182, "num_tokens": 189030706.0, "step": 4953 }, { "epoch": 0.6301997201373871, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.437722206115723, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8813658952713013, "num_tokens": 189065638.0, "step": 4954 }, { "epoch": 0.6303269304159776, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.377420425415039, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8724892139434814, "num_tokens": 189101747.0, "step": 4955 }, { "epoch": 0.6304541406945681, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.397411346435547, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.856086254119873, "num_tokens": 189142965.0, "step": 4956 }, { "epoch": 0.6305813509731586, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.4297513961792, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8646465539932251, "num_tokens": 189177822.0, "step": 4957 }, { "epoch": 0.6307085612517491, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.40371322631836, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.854072093963623, "num_tokens": 189215710.0, "step": 4958 }, { "epoch": 0.6308357715303397, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.471294403076172, "learning_rate": 1e-06, "loss": 0.5254, "mean_token_accuracy": 0.8331524729728699, "num_tokens": 189250788.0, "step": 4959 }, { "epoch": 0.6309629818089302, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.4381103515625, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8576211929321289, "num_tokens": 189278451.0, "step": 4960 }, { "epoch": 0.6310901920875207, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.435561180114746, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8629956245422363, "num_tokens": 189316857.0, "step": 4961 }, { "epoch": 0.6312174023661112, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.361404418945312, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8711904287338257, "num_tokens": 189356600.0, "step": 4962 }, { "epoch": 0.6313446126447017, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.446769714355469, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8552196025848389, "num_tokens": 189393134.0, "step": 4963 }, { "epoch": 0.6314718229232922, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.407678604125977, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.844077467918396, "num_tokens": 189433431.0, "step": 4964 }, { "epoch": 0.6315990332018827, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.386292457580566, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8468453884124756, "num_tokens": 189473738.0, "step": 4965 }, { "epoch": 0.6317262434804732, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.462407112121582, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8734110593795776, "num_tokens": 189509930.0, "step": 4966 }, { "epoch": 0.6318534537590638, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.409390449523926, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8644603490829468, "num_tokens": 189547801.0, "step": 4967 }, { "epoch": 0.6319806640376543, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.437697410583496, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8717198371887207, "num_tokens": 189582226.0, "step": 4968 }, { "epoch": 0.6321078743162447, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.399779319763184, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8586476445198059, "num_tokens": 189622915.0, "step": 4969 }, { "epoch": 0.6322350845948352, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.458329200744629, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8675436973571777, "num_tokens": 189658182.0, "step": 4970 }, { "epoch": 0.6323622948734258, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.458017349243164, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8618305921554565, "num_tokens": 189696627.0, "step": 4971 }, { "epoch": 0.6324895051520163, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.497044563293457, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8581971526145935, "num_tokens": 189729826.0, "step": 4972 }, { "epoch": 0.6326167154306068, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.397421836853027, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8647136092185974, "num_tokens": 189771504.0, "step": 4973 }, { "epoch": 0.6327439257091974, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.429865837097168, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8769996166229248, "num_tokens": 189809220.0, "step": 4974 }, { "epoch": 0.6328711359877878, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.49628734588623, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8676257729530334, "num_tokens": 189849586.0, "step": 4975 }, { "epoch": 0.6329983462663783, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.423662185668945, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8609631061553955, "num_tokens": 189885054.0, "step": 4976 }, { "epoch": 0.6331255565449688, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.44038200378418, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8531767725944519, "num_tokens": 189918286.0, "step": 4977 }, { "epoch": 0.6332527668235594, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.444009780883789, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8412978649139404, "num_tokens": 189956845.0, "step": 4978 }, { "epoch": 0.6333799771021499, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.406004905700684, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8774563074111938, "num_tokens": 189994827.0, "step": 4979 }, { "epoch": 0.6335071873807404, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.401616096496582, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8542738556861877, "num_tokens": 190036066.0, "step": 4980 }, { "epoch": 0.6336343976593308, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.414721488952637, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8709052801132202, "num_tokens": 190073326.0, "step": 4981 }, { "epoch": 0.6337616079379214, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.435789108276367, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8584412336349487, "num_tokens": 190109125.0, "step": 4982 }, { "epoch": 0.6338888182165119, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.44173812866211, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8659737706184387, "num_tokens": 190148520.0, "step": 4983 }, { "epoch": 0.6340160284951024, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.472125053405762, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8624565005302429, "num_tokens": 190184272.0, "step": 4984 }, { "epoch": 0.6341432387736929, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.492923736572266, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8597115278244019, "num_tokens": 190221483.0, "step": 4985 }, { "epoch": 0.6342704490522835, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.401631355285645, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8721631169319153, "num_tokens": 190266430.0, "step": 4986 }, { "epoch": 0.6343976593308739, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.430426597595215, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.871630847454071, "num_tokens": 190308575.0, "step": 4987 }, { "epoch": 0.6345248696094644, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.44682788848877, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.870602011680603, "num_tokens": 190344584.0, "step": 4988 }, { "epoch": 0.6346520798880549, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.443346977233887, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8682559728622437, "num_tokens": 190376497.0, "step": 4989 }, { "epoch": 0.6347792901666455, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.430355072021484, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8494168519973755, "num_tokens": 190416879.0, "step": 4990 }, { "epoch": 0.634906500445236, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.395565032958984, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8804056644439697, "num_tokens": 190454684.0, "step": 4991 }, { "epoch": 0.6350337107238265, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.386129379272461, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8527237176895142, "num_tokens": 190490524.0, "step": 4992 }, { "epoch": 0.635160921002417, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.419395446777344, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8678309917449951, "num_tokens": 190526558.0, "step": 4993 }, { "epoch": 0.6352881312810075, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.373167037963867, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8609440326690674, "num_tokens": 190571206.0, "step": 4994 }, { "epoch": 0.635415341559598, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.400367736816406, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8491085767745972, "num_tokens": 190611480.0, "step": 4995 }, { "epoch": 0.6355425518381885, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.496232986450195, "learning_rate": 1e-06, "loss": 0.5113, "mean_token_accuracy": 0.8343162536621094, "num_tokens": 190647699.0, "step": 4996 }, { "epoch": 0.6356697621167791, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.435072898864746, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8607181906700134, "num_tokens": 190684863.0, "step": 4997 }, { "epoch": 0.6357969723953696, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.398992538452148, "learning_rate": 1e-06, "loss": 0.4783, "mean_token_accuracy": 0.8462756276130676, "num_tokens": 190725006.0, "step": 4998 }, { "epoch": 0.63592418267396, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.469285011291504, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8574953079223633, "num_tokens": 190757431.0, "step": 4999 }, { "epoch": 0.6360513929525505, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.459362030029297, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8657665252685547, "num_tokens": 190793118.0, "step": 5000 }, { "epoch": 0.6361786032311411, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.404587745666504, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8700188994407654, "num_tokens": 190837105.0, "step": 5001 }, { "epoch": 0.6363058135097316, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.427102088928223, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8549068570137024, "num_tokens": 190875372.0, "step": 5002 }, { "epoch": 0.6364330237883221, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.368487358093262, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8654979467391968, "num_tokens": 190921307.0, "step": 5003 }, { "epoch": 0.6365602340669126, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.455360412597656, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8608156442642212, "num_tokens": 190967919.0, "step": 5004 }, { "epoch": 0.6366874443455031, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.455770492553711, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8596464991569519, "num_tokens": 191005668.0, "step": 5005 }, { "epoch": 0.6368146546240936, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.421022415161133, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8495628833770752, "num_tokens": 191044455.0, "step": 5006 }, { "epoch": 0.6369418649026841, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.368589401245117, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8654359579086304, "num_tokens": 191086369.0, "step": 5007 }, { "epoch": 0.6370690751812746, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.470584869384766, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8553494215011597, "num_tokens": 191127070.0, "step": 5008 }, { "epoch": 0.6371962854598652, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.468512535095215, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8745014071464539, "num_tokens": 191163623.0, "step": 5009 }, { "epoch": 0.6373234957384557, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.46890640258789, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8567800521850586, "num_tokens": 191196790.0, "step": 5010 }, { "epoch": 0.6374507060170462, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.52609920501709, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8576294183731079, "num_tokens": 191230265.0, "step": 5011 }, { "epoch": 0.6375779162956366, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.442899703979492, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8523322939872742, "num_tokens": 191266781.0, "step": 5012 }, { "epoch": 0.6377051265742272, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.434762954711914, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8503866791725159, "num_tokens": 191305869.0, "step": 5013 }, { "epoch": 0.6378323368528177, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.431828498840332, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8489962220191956, "num_tokens": 191343498.0, "step": 5014 }, { "epoch": 0.6379595471314082, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.43720531463623, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.8458915948867798, "num_tokens": 191383260.0, "step": 5015 }, { "epoch": 0.6380867574099988, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.46647834777832, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8513399958610535, "num_tokens": 191417350.0, "step": 5016 }, { "epoch": 0.6382139676885893, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.432899475097656, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8618470430374146, "num_tokens": 191454336.0, "step": 5017 }, { "epoch": 0.6383411779671797, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.455366134643555, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8614046573638916, "num_tokens": 191489977.0, "step": 5018 }, { "epoch": 0.6384683882457702, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.461540222167969, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8507587909698486, "num_tokens": 191524622.0, "step": 5019 }, { "epoch": 0.6385955985243608, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.440199851989746, "learning_rate": 1e-06, "loss": 0.4649, "mean_token_accuracy": 0.8489493131637573, "num_tokens": 191565389.0, "step": 5020 }, { "epoch": 0.6387228088029513, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.422701835632324, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8795472383499146, "num_tokens": 191604640.0, "step": 5021 }, { "epoch": 0.6388500190815418, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.407632827758789, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.859499990940094, "num_tokens": 191651029.0, "step": 5022 }, { "epoch": 0.6389772293601323, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.462430953979492, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8605583906173706, "num_tokens": 191690524.0, "step": 5023 }, { "epoch": 0.6391044396387228, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.408994674682617, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8555098176002502, "num_tokens": 191733992.0, "step": 5024 }, { "epoch": 0.6392316499173133, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.408130645751953, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8706180453300476, "num_tokens": 191769717.0, "step": 5025 }, { "epoch": 0.6393588601959038, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.456607818603516, "learning_rate": 1e-06, "loss": 0.4807, "mean_token_accuracy": 0.8456310033798218, "num_tokens": 191813334.0, "step": 5026 }, { "epoch": 0.6394860704744944, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.482550621032715, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8619289994239807, "num_tokens": 191848295.0, "step": 5027 }, { "epoch": 0.6396132807530849, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.456811904907227, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8531159162521362, "num_tokens": 191885329.0, "step": 5028 }, { "epoch": 0.6397404910316754, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.452112197875977, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8732640743255615, "num_tokens": 191918070.0, "step": 5029 }, { "epoch": 0.6398677013102658, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.489380836486816, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8639468550682068, "num_tokens": 191966084.0, "step": 5030 }, { "epoch": 0.6399949115888564, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.46718692779541, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8595454692840576, "num_tokens": 192004907.0, "step": 5031 }, { "epoch": 0.6401221218674469, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.50199031829834, "learning_rate": 1e-06, "loss": 0.4872, "mean_token_accuracy": 0.841606616973877, "num_tokens": 192042942.0, "step": 5032 }, { "epoch": 0.6402493321460374, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.454458236694336, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8603315949440002, "num_tokens": 192081133.0, "step": 5033 }, { "epoch": 0.6403765424246279, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.449140548706055, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.860303521156311, "num_tokens": 192120171.0, "step": 5034 }, { "epoch": 0.6405037527032185, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.522734642028809, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8606092929840088, "num_tokens": 192159583.0, "step": 5035 }, { "epoch": 0.6406309629818089, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.500965118408203, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8583850860595703, "num_tokens": 192199513.0, "step": 5036 }, { "epoch": 0.6407581732603994, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.479788780212402, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8625720739364624, "num_tokens": 192228352.0, "step": 5037 }, { "epoch": 0.6408853835389899, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.490997314453125, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.855527937412262, "num_tokens": 192264480.0, "step": 5038 }, { "epoch": 0.6410125938175805, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.49950885772705, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8583667278289795, "num_tokens": 192303307.0, "step": 5039 }, { "epoch": 0.641139804096171, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.495404243469238, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8631489872932434, "num_tokens": 192342064.0, "step": 5040 }, { "epoch": 0.6412670143747615, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.459671974182129, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8680112361907959, "num_tokens": 192379297.0, "step": 5041 }, { "epoch": 0.6413942246533519, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.504828453063965, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8635388612747192, "num_tokens": 192416087.0, "step": 5042 }, { "epoch": 0.6415214349319425, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.419013023376465, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8657046556472778, "num_tokens": 192454421.0, "step": 5043 }, { "epoch": 0.641648645210533, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.468552589416504, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8572126626968384, "num_tokens": 192493860.0, "step": 5044 }, { "epoch": 0.6417758554891235, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.404534339904785, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8518683910369873, "num_tokens": 192533672.0, "step": 5045 }, { "epoch": 0.641903065767714, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.50788688659668, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8479993343353271, "num_tokens": 192569569.0, "step": 5046 }, { "epoch": 0.6420302760463046, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.479286193847656, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8661242723464966, "num_tokens": 192610716.0, "step": 5047 }, { "epoch": 0.642157486324895, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.44257640838623, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8540279865264893, "num_tokens": 192648040.0, "step": 5048 }, { "epoch": 0.6422846966034855, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.480135917663574, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8587084412574768, "num_tokens": 192690063.0, "step": 5049 }, { "epoch": 0.6424119068820761, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.426421165466309, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8556678295135498, "num_tokens": 192732067.0, "step": 5050 }, { "epoch": 0.6425391171606666, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.50149917602539, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8702239990234375, "num_tokens": 192768598.0, "step": 5051 }, { "epoch": 0.6426663274392571, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.445090293884277, "learning_rate": 1e-06, "loss": 0.5246, "mean_token_accuracy": 0.8326868414878845, "num_tokens": 192805420.0, "step": 5052 }, { "epoch": 0.6427935377178476, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.484848976135254, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8557671904563904, "num_tokens": 192852082.0, "step": 5053 }, { "epoch": 0.6429207479964381, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.402507781982422, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8612262606620789, "num_tokens": 192892846.0, "step": 5054 }, { "epoch": 0.6430479582750286, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.48248291015625, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8780114650726318, "num_tokens": 192932091.0, "step": 5055 }, { "epoch": 0.6431751685536191, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.406878471374512, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8619045615196228, "num_tokens": 192960628.0, "step": 5056 }, { "epoch": 0.6433023788322096, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.472543716430664, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8533105850219727, "num_tokens": 193002075.0, "step": 5057 }, { "epoch": 0.6434295891108002, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.405049324035645, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8489834070205688, "num_tokens": 193039720.0, "step": 5058 }, { "epoch": 0.6435567993893907, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.544074058532715, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8597761392593384, "num_tokens": 193072806.0, "step": 5059 }, { "epoch": 0.6436840096679812, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.48984432220459, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8694688081741333, "num_tokens": 193108478.0, "step": 5060 }, { "epoch": 0.6438112199465716, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.477983474731445, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8638186454772949, "num_tokens": 193150663.0, "step": 5061 }, { "epoch": 0.6439384302251622, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.440811157226562, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8633713126182556, "num_tokens": 193191499.0, "step": 5062 }, { "epoch": 0.6440656405037527, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.459766387939453, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.85089111328125, "num_tokens": 193236103.0, "step": 5063 }, { "epoch": 0.6441928507823432, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.494791030883789, "learning_rate": 1e-06, "loss": 0.5031, "mean_token_accuracy": 0.8387176990509033, "num_tokens": 193274095.0, "step": 5064 }, { "epoch": 0.6443200610609338, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.434561729431152, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8659564256668091, "num_tokens": 193316988.0, "step": 5065 }, { "epoch": 0.6444472713395243, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.468989372253418, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.858744740486145, "num_tokens": 193357430.0, "step": 5066 }, { "epoch": 0.6445744816181147, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.509765625, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8680835962295532, "num_tokens": 193390392.0, "step": 5067 }, { "epoch": 0.6447016918967052, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.430041313171387, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8577354550361633, "num_tokens": 193430961.0, "step": 5068 }, { "epoch": 0.6448289021752958, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.552632331848145, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8584866523742676, "num_tokens": 193473767.0, "step": 5069 }, { "epoch": 0.6449561124538863, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.521693229675293, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8487505316734314, "num_tokens": 193511375.0, "step": 5070 }, { "epoch": 0.6450833227324768, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.496391296386719, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8703407049179077, "num_tokens": 193551296.0, "step": 5071 }, { "epoch": 0.6452105330110673, "ewc_loss": 0.01904296875, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.43897533416748, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8636276721954346, "num_tokens": 193596484.0, "step": 5072 }, { "epoch": 0.6453377432896578, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.490816116333008, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8663505911827087, "num_tokens": 193637513.0, "step": 5073 }, { "epoch": 0.6454649535682483, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.456825256347656, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8543897271156311, "num_tokens": 193672808.0, "step": 5074 }, { "epoch": 0.6455921638468388, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.442782402038574, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.875126838684082, "num_tokens": 193713018.0, "step": 5075 }, { "epoch": 0.6457193741254293, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.443937301635742, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8617156147956848, "num_tokens": 193757649.0, "step": 5076 }, { "epoch": 0.6458465844040199, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.499894142150879, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8567202687263489, "num_tokens": 193792390.0, "step": 5077 }, { "epoch": 0.6459737946826104, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.45402717590332, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8671852946281433, "num_tokens": 193827721.0, "step": 5078 }, { "epoch": 0.6461010049612008, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.48004150390625, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8517075777053833, "num_tokens": 193867217.0, "step": 5079 }, { "epoch": 0.6462282152397913, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.491166114807129, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.868035078048706, "num_tokens": 193900128.0, "step": 5080 }, { "epoch": 0.6463554255183819, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.50726318359375, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8700852394104004, "num_tokens": 193935744.0, "step": 5081 }, { "epoch": 0.6464826357969724, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.438248634338379, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8575423359870911, "num_tokens": 193977344.0, "step": 5082 }, { "epoch": 0.6466098460755629, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.490564346313477, "learning_rate": 1e-06, "loss": 0.5168, "mean_token_accuracy": 0.8346315026283264, "num_tokens": 194014295.0, "step": 5083 }, { "epoch": 0.6467370563541535, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.401183128356934, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8449744582176208, "num_tokens": 194055245.0, "step": 5084 }, { "epoch": 0.6468642666327439, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.444705963134766, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8528525829315186, "num_tokens": 194089313.0, "step": 5085 }, { "epoch": 0.6469914769113344, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.488741874694824, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.865987241268158, "num_tokens": 194128293.0, "step": 5086 }, { "epoch": 0.6471186871899249, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.509065628051758, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.853549599647522, "num_tokens": 194166632.0, "step": 5087 }, { "epoch": 0.6472458974685155, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.456918716430664, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8460656404495239, "num_tokens": 194209176.0, "step": 5088 }, { "epoch": 0.647373107747106, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.527685165405273, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8499789834022522, "num_tokens": 194245617.0, "step": 5089 }, { "epoch": 0.6475003180256965, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.449493408203125, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8664770126342773, "num_tokens": 194285983.0, "step": 5090 }, { "epoch": 0.6476275283042869, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.481143951416016, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8560276031494141, "num_tokens": 194317088.0, "step": 5091 }, { "epoch": 0.6477547385828775, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.490456581115723, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.86448073387146, "num_tokens": 194354025.0, "step": 5092 }, { "epoch": 0.647881948861468, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.424545288085938, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8636256456375122, "num_tokens": 194388539.0, "step": 5093 }, { "epoch": 0.6480091591400585, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.520781517028809, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8622636198997498, "num_tokens": 194424518.0, "step": 5094 }, { "epoch": 0.648136369418649, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.398661613464355, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8566685914993286, "num_tokens": 194463310.0, "step": 5095 }, { "epoch": 0.6482635796972396, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.47616195678711, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8596029877662659, "num_tokens": 194499911.0, "step": 5096 }, { "epoch": 0.64839078997583, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.508101463317871, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8503124117851257, "num_tokens": 194536731.0, "step": 5097 }, { "epoch": 0.6485180002544205, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.459839820861816, "learning_rate": 1e-06, "loss": 0.4721, "mean_token_accuracy": 0.8497669100761414, "num_tokens": 194574014.0, "step": 5098 }, { "epoch": 0.648645210533011, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.516928672790527, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8515995144844055, "num_tokens": 194611944.0, "step": 5099 }, { "epoch": 0.6487724208116016, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.491863250732422, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8693551421165466, "num_tokens": 194649734.0, "step": 5100 }, { "epoch": 0.6488996310901921, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.505797386169434, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8680945634841919, "num_tokens": 194688344.0, "step": 5101 }, { "epoch": 0.6490268413687826, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.502745628356934, "learning_rate": 1e-06, "loss": 0.4926, "mean_token_accuracy": 0.8470761775970459, "num_tokens": 194728582.0, "step": 5102 }, { "epoch": 0.649154051647373, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.41796875, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8593893051147461, "num_tokens": 194762753.0, "step": 5103 }, { "epoch": 0.6492812619259636, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.449198722839355, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8730789422988892, "num_tokens": 194802325.0, "step": 5104 }, { "epoch": 0.6494084722045541, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.384344100952148, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8587691187858582, "num_tokens": 194846349.0, "step": 5105 }, { "epoch": 0.6495356824831446, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.441495895385742, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.860327959060669, "num_tokens": 194880334.0, "step": 5106 }, { "epoch": 0.6496628927617352, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.475861549377441, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8508855700492859, "num_tokens": 194919343.0, "step": 5107 }, { "epoch": 0.6497901030403257, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.472070693969727, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8605560064315796, "num_tokens": 194958800.0, "step": 5108 }, { "epoch": 0.6499173133189162, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.41664981842041, "learning_rate": 1e-06, "loss": 0.4995, "mean_token_accuracy": 0.832964301109314, "num_tokens": 194997249.0, "step": 5109 }, { "epoch": 0.6500445235975066, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.49252986907959, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8616476058959961, "num_tokens": 195035879.0, "step": 5110 }, { "epoch": 0.6501717338760972, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.429637908935547, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8547728657722473, "num_tokens": 195075925.0, "step": 5111 }, { "epoch": 0.6502989441546877, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.460155487060547, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8623016476631165, "num_tokens": 195110185.0, "step": 5112 }, { "epoch": 0.6504261544332782, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.443870544433594, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8513493537902832, "num_tokens": 195154666.0, "step": 5113 }, { "epoch": 0.6505533647118688, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.454926490783691, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8716210126876831, "num_tokens": 195188998.0, "step": 5114 }, { "epoch": 0.6506805749904593, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.50765323638916, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8672019243240356, "num_tokens": 195224124.0, "step": 5115 }, { "epoch": 0.6508077852690497, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.494994163513184, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8595256805419922, "num_tokens": 195253601.0, "step": 5116 }, { "epoch": 0.6509349955476402, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.525528907775879, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8511689901351929, "num_tokens": 195289045.0, "step": 5117 }, { "epoch": 0.6510622058262308, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.508155822753906, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8619294762611389, "num_tokens": 195326983.0, "step": 5118 }, { "epoch": 0.6511894161048213, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.4808349609375, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8720301389694214, "num_tokens": 195360425.0, "step": 5119 }, { "epoch": 0.6513166263834118, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.504528045654297, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.848808765411377, "num_tokens": 195402648.0, "step": 5120 }, { "epoch": 0.6514438366620023, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.493096351623535, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8777273893356323, "num_tokens": 195440945.0, "step": 5121 }, { "epoch": 0.6515710469405928, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.497407913208008, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8621369004249573, "num_tokens": 195481872.0, "step": 5122 }, { "epoch": 0.6516982572191833, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.460810661315918, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8642977476119995, "num_tokens": 195525460.0, "step": 5123 }, { "epoch": 0.6518254674977738, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.477434158325195, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8618996143341064, "num_tokens": 195563522.0, "step": 5124 }, { "epoch": 0.6519526777763643, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.532177925109863, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8696385622024536, "num_tokens": 195598114.0, "step": 5125 }, { "epoch": 0.6520798880549549, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.504172325134277, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8548825979232788, "num_tokens": 195639826.0, "step": 5126 }, { "epoch": 0.6522070983335454, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.458420753479004, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8668844699859619, "num_tokens": 195682047.0, "step": 5127 }, { "epoch": 0.6523343086121358, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.473137855529785, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8647654056549072, "num_tokens": 195722453.0, "step": 5128 }, { "epoch": 0.6524615188907263, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.497032165527344, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8552803993225098, "num_tokens": 195759182.0, "step": 5129 }, { "epoch": 0.6525887291693169, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.507603645324707, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8634878396987915, "num_tokens": 195795641.0, "step": 5130 }, { "epoch": 0.6527159394479074, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.525176048278809, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8700657486915588, "num_tokens": 195831200.0, "step": 5131 }, { "epoch": 0.6528431497264979, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.475605964660645, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.863169252872467, "num_tokens": 195875110.0, "step": 5132 }, { "epoch": 0.6529703600050885, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.510534286499023, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8631308078765869, "num_tokens": 195916724.0, "step": 5133 }, { "epoch": 0.6530975702836789, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.544639587402344, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8547205924987793, "num_tokens": 195956219.0, "step": 5134 }, { "epoch": 0.6532247805622694, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.523283004760742, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8640199899673462, "num_tokens": 195991563.0, "step": 5135 }, { "epoch": 0.6533519908408599, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.53560733795166, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8675374388694763, "num_tokens": 196031864.0, "step": 5136 }, { "epoch": 0.6534792011194505, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.531752586364746, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8523671627044678, "num_tokens": 196067208.0, "step": 5137 }, { "epoch": 0.653606411398041, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.525674819946289, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.863233745098114, "num_tokens": 196103934.0, "step": 5138 }, { "epoch": 0.6537336216766315, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.526627540588379, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8553100824356079, "num_tokens": 196140886.0, "step": 5139 }, { "epoch": 0.6538608319552219, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.508017539978027, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8695849180221558, "num_tokens": 196178940.0, "step": 5140 }, { "epoch": 0.6539880422338125, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.521954536437988, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.862173318862915, "num_tokens": 196229461.0, "step": 5141 }, { "epoch": 0.654115252512403, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.499207496643066, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8735107779502869, "num_tokens": 196265793.0, "step": 5142 }, { "epoch": 0.6542424627909935, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.453969955444336, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.867429256439209, "num_tokens": 196304919.0, "step": 5143 }, { "epoch": 0.654369673069584, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.528854370117188, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8756413459777832, "num_tokens": 196344761.0, "step": 5144 }, { "epoch": 0.6544968833481746, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.46934700012207, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8750810623168945, "num_tokens": 196382625.0, "step": 5145 }, { "epoch": 0.654624093626765, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.459460258483887, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8722620010375977, "num_tokens": 196422279.0, "step": 5146 }, { "epoch": 0.6547513039053555, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.480090141296387, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.856661319732666, "num_tokens": 196463684.0, "step": 5147 }, { "epoch": 0.654878514183946, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.445136070251465, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8634698987007141, "num_tokens": 196506319.0, "step": 5148 }, { "epoch": 0.6550057244625366, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.47191333770752, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.854314923286438, "num_tokens": 196546367.0, "step": 5149 }, { "epoch": 0.6551329347411271, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.482808113098145, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8464201092720032, "num_tokens": 196576598.0, "step": 5150 }, { "epoch": 0.6552601450197176, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.485766410827637, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8546360731124878, "num_tokens": 196610633.0, "step": 5151 }, { "epoch": 0.655387355298308, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.542147636413574, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8514443635940552, "num_tokens": 196646183.0, "step": 5152 }, { "epoch": 0.6555145655768986, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.435885429382324, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.869961142539978, "num_tokens": 196689454.0, "step": 5153 }, { "epoch": 0.6556417758554891, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.444181442260742, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.8456281423568726, "num_tokens": 196731820.0, "step": 5154 }, { "epoch": 0.6557689861340796, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.542017936706543, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8509939908981323, "num_tokens": 196768542.0, "step": 5155 }, { "epoch": 0.6558961964126702, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.455626487731934, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.845773458480835, "num_tokens": 196810300.0, "step": 5156 }, { "epoch": 0.6560234066912607, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.473620414733887, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8443684577941895, "num_tokens": 196847271.0, "step": 5157 }, { "epoch": 0.6561506169698512, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.434237480163574, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8752429485321045, "num_tokens": 196880439.0, "step": 5158 }, { "epoch": 0.6562778272484416, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.494736671447754, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8532203435897827, "num_tokens": 196914911.0, "step": 5159 }, { "epoch": 0.6564050375270322, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.513628005981445, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.860980749130249, "num_tokens": 196949419.0, "step": 5160 }, { "epoch": 0.6565322478056227, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.51036548614502, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8528019785881042, "num_tokens": 196989401.0, "step": 5161 }, { "epoch": 0.6566594580842132, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.46645450592041, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8533873558044434, "num_tokens": 197028377.0, "step": 5162 }, { "epoch": 0.6567866683628037, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.552017211914062, "learning_rate": 1e-06, "loss": 0.4932, "mean_token_accuracy": 0.8403285145759583, "num_tokens": 197068334.0, "step": 5163 }, { "epoch": 0.6569138786413943, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.445453643798828, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8629382848739624, "num_tokens": 197105978.0, "step": 5164 }, { "epoch": 0.6570410889199847, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.499865531921387, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8433468341827393, "num_tokens": 197148641.0, "step": 5165 }, { "epoch": 0.6571682991985752, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.480720520019531, "learning_rate": 1e-06, "loss": 0.4966, "mean_token_accuracy": 0.8414371013641357, "num_tokens": 197188820.0, "step": 5166 }, { "epoch": 0.6572955094771658, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.467158317565918, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8476989269256592, "num_tokens": 197227109.0, "step": 5167 }, { "epoch": 0.6574227197557563, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.515987396240234, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8657155632972717, "num_tokens": 197264083.0, "step": 5168 }, { "epoch": 0.6575499300343468, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.473692893981934, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8561704158782959, "num_tokens": 197301349.0, "step": 5169 }, { "epoch": 0.6576771403129373, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.508367538452148, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8496230840682983, "num_tokens": 197348176.0, "step": 5170 }, { "epoch": 0.6578043505915278, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.549088478088379, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8524652719497681, "num_tokens": 197380490.0, "step": 5171 }, { "epoch": 0.6579315608701183, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.46993637084961, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.846467137336731, "num_tokens": 197423075.0, "step": 5172 }, { "epoch": 0.6580587711487088, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.524669647216797, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8646290898323059, "num_tokens": 197461730.0, "step": 5173 }, { "epoch": 0.6581859814272993, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.521864891052246, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8489557504653931, "num_tokens": 197505801.0, "step": 5174 }, { "epoch": 0.6583131917058899, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.52912425994873, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8769787549972534, "num_tokens": 197534643.0, "step": 5175 }, { "epoch": 0.6584404019844804, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.440044403076172, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8660818338394165, "num_tokens": 197576490.0, "step": 5176 }, { "epoch": 0.6585676122630708, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.524947166442871, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8736517429351807, "num_tokens": 197612745.0, "step": 5177 }, { "epoch": 0.6586948225416613, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.453018188476562, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8555880188941956, "num_tokens": 197653756.0, "step": 5178 }, { "epoch": 0.6588220328202519, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.511066436767578, "learning_rate": 1e-06, "loss": 0.3525, "mean_token_accuracy": 0.884394645690918, "num_tokens": 197694417.0, "step": 5179 }, { "epoch": 0.6589492430988424, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.475465774536133, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8556049466133118, "num_tokens": 197732738.0, "step": 5180 }, { "epoch": 0.6590764533774329, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.551814079284668, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8513413667678833, "num_tokens": 197759408.0, "step": 5181 }, { "epoch": 0.6592036636560235, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.48526668548584, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8620616793632507, "num_tokens": 197794334.0, "step": 5182 }, { "epoch": 0.6593308739346139, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.47622013092041, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8689601421356201, "num_tokens": 197832497.0, "step": 5183 }, { "epoch": 0.6594580842132044, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.493078231811523, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.851874589920044, "num_tokens": 197869021.0, "step": 5184 }, { "epoch": 0.6595852944917949, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.516315460205078, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8463060259819031, "num_tokens": 197906365.0, "step": 5185 }, { "epoch": 0.6597125047703855, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.431556701660156, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8629530072212219, "num_tokens": 197950725.0, "step": 5186 }, { "epoch": 0.659839715048976, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.468902587890625, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8680094480514526, "num_tokens": 197991178.0, "step": 5187 }, { "epoch": 0.6599669253275665, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.491621017456055, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8514808416366577, "num_tokens": 198030176.0, "step": 5188 }, { "epoch": 0.6600941356061569, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.532718658447266, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8543437719345093, "num_tokens": 198067169.0, "step": 5189 }, { "epoch": 0.6602213458847475, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.476799011230469, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8491774201393127, "num_tokens": 198104994.0, "step": 5190 }, { "epoch": 0.660348556163338, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.461280822753906, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8657048344612122, "num_tokens": 198142791.0, "step": 5191 }, { "epoch": 0.6604757664419285, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.501872062683105, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8578652143478394, "num_tokens": 198179300.0, "step": 5192 }, { "epoch": 0.660602976720519, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.469508171081543, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8744189143180847, "num_tokens": 198220230.0, "step": 5193 }, { "epoch": 0.6607301869991096, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.594398498535156, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8483762145042419, "num_tokens": 198251067.0, "step": 5194 }, { "epoch": 0.6608573972777, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.519410133361816, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8687145709991455, "num_tokens": 198282130.0, "step": 5195 }, { "epoch": 0.6609846075562905, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.453628540039062, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.869701087474823, "num_tokens": 198325609.0, "step": 5196 }, { "epoch": 0.661111817834881, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.516338348388672, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8638101816177368, "num_tokens": 198365116.0, "step": 5197 }, { "epoch": 0.6612390281134716, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.50727653503418, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8606869578361511, "num_tokens": 198403467.0, "step": 5198 }, { "epoch": 0.6613662383920621, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.519222259521484, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8664131164550781, "num_tokens": 198440676.0, "step": 5199 }, { "epoch": 0.6614934486706526, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.5107421875, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8624531626701355, "num_tokens": 198479082.0, "step": 5200 }, { "epoch": 0.661620658949243, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.4924955368042, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8576985597610474, "num_tokens": 198521741.0, "step": 5201 }, { "epoch": 0.6617478692278336, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.519536018371582, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8541151881217957, "num_tokens": 198562177.0, "step": 5202 }, { "epoch": 0.6618750795064241, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.53367805480957, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8614318370819092, "num_tokens": 198600726.0, "step": 5203 }, { "epoch": 0.6620022897850146, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.593008995056152, "learning_rate": 1e-06, "loss": 0.5224, "mean_token_accuracy": 0.8312684297561646, "num_tokens": 198638156.0, "step": 5204 }, { "epoch": 0.6621295000636052, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.523286819458008, "learning_rate": 1e-06, "loss": 0.5074, "mean_token_accuracy": 0.8411329984664917, "num_tokens": 198680312.0, "step": 5205 }, { "epoch": 0.6622567103421957, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.468236923217773, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8781524896621704, "num_tokens": 198723886.0, "step": 5206 }, { "epoch": 0.6623839206207861, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.53699016571045, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8597793579101562, "num_tokens": 198765331.0, "step": 5207 }, { "epoch": 0.6625111308993766, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.536383628845215, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.864448606967926, "num_tokens": 198802421.0, "step": 5208 }, { "epoch": 0.6626383411779672, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.48997974395752, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8701942563056946, "num_tokens": 198843335.0, "step": 5209 }, { "epoch": 0.6627655514565577, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.519891738891602, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8671237230300903, "num_tokens": 198880925.0, "step": 5210 }, { "epoch": 0.6628927617351482, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.583388328552246, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8596948385238647, "num_tokens": 198915448.0, "step": 5211 }, { "epoch": 0.6630199720137387, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.489749908447266, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8709497451782227, "num_tokens": 198952254.0, "step": 5212 }, { "epoch": 0.6631471822923293, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.5991792678833, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8653099536895752, "num_tokens": 198990841.0, "step": 5213 }, { "epoch": 0.6632743925709197, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.533348083496094, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8636941909790039, "num_tokens": 199029383.0, "step": 5214 }, { "epoch": 0.6634016028495102, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.500329971313477, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8598905205726624, "num_tokens": 199066731.0, "step": 5215 }, { "epoch": 0.6635288131281007, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.505473136901855, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8725875616073608, "num_tokens": 199108752.0, "step": 5216 }, { "epoch": 0.6636560234066913, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.566198348999023, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.86405348777771, "num_tokens": 199146438.0, "step": 5217 }, { "epoch": 0.6637832336852818, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.501067161560059, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8627375364303589, "num_tokens": 199186367.0, "step": 5218 }, { "epoch": 0.6639104439638723, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.49592399597168, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8745390176773071, "num_tokens": 199225785.0, "step": 5219 }, { "epoch": 0.6640376542424627, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.53821849822998, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8637948036193848, "num_tokens": 199262574.0, "step": 5220 }, { "epoch": 0.6641648645210533, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.553727149963379, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8668403625488281, "num_tokens": 199299093.0, "step": 5221 }, { "epoch": 0.6642920747996438, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.549166679382324, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.866614580154419, "num_tokens": 199337446.0, "step": 5222 }, { "epoch": 0.6644192850782343, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.507126808166504, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8627214431762695, "num_tokens": 199377321.0, "step": 5223 }, { "epoch": 0.6645464953568249, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.55255126953125, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8713725805282593, "num_tokens": 199413057.0, "step": 5224 }, { "epoch": 0.6646737056354154, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.565424919128418, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8676928877830505, "num_tokens": 199452730.0, "step": 5225 }, { "epoch": 0.6648009159140058, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.480916976928711, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.871125340461731, "num_tokens": 199487860.0, "step": 5226 }, { "epoch": 0.6649281261925963, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.506196975708008, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8678569793701172, "num_tokens": 199532790.0, "step": 5227 }, { "epoch": 0.6650553364711869, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.56476879119873, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8814600110054016, "num_tokens": 199561072.0, "step": 5228 }, { "epoch": 0.6651825467497774, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.44077205657959, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8673135042190552, "num_tokens": 199595820.0, "step": 5229 }, { "epoch": 0.6653097570283679, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.516246795654297, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8489252328872681, "num_tokens": 199640101.0, "step": 5230 }, { "epoch": 0.6654369673069584, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.452059745788574, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8579787611961365, "num_tokens": 199680287.0, "step": 5231 }, { "epoch": 0.6655641775855489, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.445015907287598, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8600598573684692, "num_tokens": 199721958.0, "step": 5232 }, { "epoch": 0.6656913878641394, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.571222305297852, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8673669099807739, "num_tokens": 199756021.0, "step": 5233 }, { "epoch": 0.6658185981427299, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.551974296569824, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8657225370407104, "num_tokens": 199793586.0, "step": 5234 }, { "epoch": 0.6659458084213205, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.482447624206543, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8481218814849854, "num_tokens": 199835728.0, "step": 5235 }, { "epoch": 0.666073018699911, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.469831466674805, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.870690643787384, "num_tokens": 199873348.0, "step": 5236 }, { "epoch": 0.6662002289785015, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.513885498046875, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8613760471343994, "num_tokens": 199908223.0, "step": 5237 }, { "epoch": 0.6663274392570919, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.552874565124512, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.857932448387146, "num_tokens": 199947017.0, "step": 5238 }, { "epoch": 0.6664546495356825, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.593255996704102, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8702240586280823, "num_tokens": 199984806.0, "step": 5239 }, { "epoch": 0.666581859814273, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.533985137939453, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8554893732070923, "num_tokens": 200022600.0, "step": 5240 }, { "epoch": 0.6667090700928635, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.537686347961426, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8716932535171509, "num_tokens": 200058407.0, "step": 5241 }, { "epoch": 0.666836280371454, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.534608840942383, "learning_rate": 1e-06, "loss": 0.5126, "mean_token_accuracy": 0.83460533618927, "num_tokens": 200101871.0, "step": 5242 }, { "epoch": 0.6669634906500446, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.506059646606445, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.862361490726471, "num_tokens": 200138501.0, "step": 5243 }, { "epoch": 0.667090700928635, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.506723403930664, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8598005771636963, "num_tokens": 200176639.0, "step": 5244 }, { "epoch": 0.6672179112072255, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.573458671569824, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8574001789093018, "num_tokens": 200211873.0, "step": 5245 }, { "epoch": 0.667345121485816, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.579645156860352, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8572925329208374, "num_tokens": 200246596.0, "step": 5246 }, { "epoch": 0.6674723317644066, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.576264381408691, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8767093420028687, "num_tokens": 200282761.0, "step": 5247 }, { "epoch": 0.6675995420429971, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.559532165527344, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8473324775695801, "num_tokens": 200321821.0, "step": 5248 }, { "epoch": 0.6677267523215876, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.571276664733887, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.844728946685791, "num_tokens": 200369066.0, "step": 5249 }, { "epoch": 0.667853962600178, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.531832695007324, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8537634015083313, "num_tokens": 200408067.0, "step": 5250 }, { "epoch": 0.6679811728787686, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.558001518249512, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8783564567565918, "num_tokens": 200446693.0, "step": 5251 }, { "epoch": 0.6681083831573591, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.542536735534668, "learning_rate": 1e-06, "loss": 0.497, "mean_token_accuracy": 0.8428008556365967, "num_tokens": 200485340.0, "step": 5252 }, { "epoch": 0.6682355934359496, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.524920463562012, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8577717542648315, "num_tokens": 200521978.0, "step": 5253 }, { "epoch": 0.6683628037145402, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.507962226867676, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8632127046585083, "num_tokens": 200560604.0, "step": 5254 }, { "epoch": 0.6684900139931307, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.546321868896484, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8568950891494751, "num_tokens": 200601718.0, "step": 5255 }, { "epoch": 0.6686172242717211, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.589699745178223, "learning_rate": 1e-06, "loss": 0.5049, "mean_token_accuracy": 0.8369619250297546, "num_tokens": 200646186.0, "step": 5256 }, { "epoch": 0.6687444345503116, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.506343841552734, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8636289834976196, "num_tokens": 200684849.0, "step": 5257 }, { "epoch": 0.6688716448289022, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.566490173339844, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8672026991844177, "num_tokens": 200719542.0, "step": 5258 }, { "epoch": 0.6689988551074927, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.530086517333984, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8621904850006104, "num_tokens": 200756212.0, "step": 5259 }, { "epoch": 0.6691260653860832, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.535737037658691, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.864033043384552, "num_tokens": 200791184.0, "step": 5260 }, { "epoch": 0.6692532756646737, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.552458763122559, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8446020483970642, "num_tokens": 200823966.0, "step": 5261 }, { "epoch": 0.6693804859432643, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.51916790008545, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8658897280693054, "num_tokens": 200862718.0, "step": 5262 }, { "epoch": 0.6695076962218547, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.539047241210938, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8565804958343506, "num_tokens": 200901755.0, "step": 5263 }, { "epoch": 0.6696349065004452, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.552459716796875, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8720008134841919, "num_tokens": 200935788.0, "step": 5264 }, { "epoch": 0.6697621167790357, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.587409019470215, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8756530284881592, "num_tokens": 200975159.0, "step": 5265 }, { "epoch": 0.6698893270576263, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.57840347290039, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8732379674911499, "num_tokens": 201010813.0, "step": 5266 }, { "epoch": 0.6700165373362168, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.54113483428955, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.859150230884552, "num_tokens": 201051937.0, "step": 5267 }, { "epoch": 0.6701437476148073, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.594878196716309, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8673567771911621, "num_tokens": 201090926.0, "step": 5268 }, { "epoch": 0.6702709578933977, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.5331449508667, "learning_rate": 1e-06, "loss": 0.4894, "mean_token_accuracy": 0.8421924710273743, "num_tokens": 201126086.0, "step": 5269 }, { "epoch": 0.6703981681719883, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.57252311706543, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8513633608818054, "num_tokens": 201165933.0, "step": 5270 }, { "epoch": 0.6705253784505788, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.576393127441406, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8641165494918823, "num_tokens": 201198794.0, "step": 5271 }, { "epoch": 0.6706525887291693, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.549012184143066, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8521929979324341, "num_tokens": 201234081.0, "step": 5272 }, { "epoch": 0.6707797990077599, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.54180908203125, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8576699495315552, "num_tokens": 201271796.0, "step": 5273 }, { "epoch": 0.6709070092863504, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.57945442199707, "learning_rate": 1e-06, "loss": 0.482, "mean_token_accuracy": 0.8429473638534546, "num_tokens": 201309526.0, "step": 5274 }, { "epoch": 0.6710342195649408, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.54311466217041, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8681620359420776, "num_tokens": 201344015.0, "step": 5275 }, { "epoch": 0.6711614298435313, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.526937484741211, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.86174476146698, "num_tokens": 201385201.0, "step": 5276 }, { "epoch": 0.6712886401221219, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.536069869995117, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.866753339767456, "num_tokens": 201422370.0, "step": 5277 }, { "epoch": 0.6714158504007124, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.549439430236816, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8466483354568481, "num_tokens": 201467085.0, "step": 5278 }, { "epoch": 0.6715430606793029, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.599008560180664, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8750421404838562, "num_tokens": 201503432.0, "step": 5279 }, { "epoch": 0.6716702709578934, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.538981437683105, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8676353693008423, "num_tokens": 201541358.0, "step": 5280 }, { "epoch": 0.6717974812364839, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.599984169006348, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8583635091781616, "num_tokens": 201580235.0, "step": 5281 }, { "epoch": 0.6719246915150744, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.5106782913208, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8614619374275208, "num_tokens": 201627703.0, "step": 5282 }, { "epoch": 0.6720519017936649, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.606344223022461, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8635272979736328, "num_tokens": 201667182.0, "step": 5283 }, { "epoch": 0.6721791120722554, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.544197082519531, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8624604344367981, "num_tokens": 201699876.0, "step": 5284 }, { "epoch": 0.672306322350846, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.557635307312012, "learning_rate": 1e-06, "loss": 0.4672, "mean_token_accuracy": 0.8522822856903076, "num_tokens": 201732770.0, "step": 5285 }, { "epoch": 0.6724335326294365, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.579139709472656, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8628515005111694, "num_tokens": 201771388.0, "step": 5286 }, { "epoch": 0.6725607429080269, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.549854278564453, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8553478121757507, "num_tokens": 201809283.0, "step": 5287 }, { "epoch": 0.6726879531866174, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.49646282196045, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8675726056098938, "num_tokens": 201854950.0, "step": 5288 }, { "epoch": 0.672815163465208, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.546857833862305, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.864841878414154, "num_tokens": 201895638.0, "step": 5289 }, { "epoch": 0.6729423737437985, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.626910209655762, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8476316332817078, "num_tokens": 201928243.0, "step": 5290 }, { "epoch": 0.673069584022389, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.522269248962402, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8611292839050293, "num_tokens": 201964687.0, "step": 5291 }, { "epoch": 0.6731967943009796, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.55773639678955, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8746017217636108, "num_tokens": 201997166.0, "step": 5292 }, { "epoch": 0.67332400457957, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.557312965393066, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8637814521789551, "num_tokens": 202030771.0, "step": 5293 }, { "epoch": 0.6734512148581605, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.50664234161377, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8534188866615295, "num_tokens": 202077280.0, "step": 5294 }, { "epoch": 0.673578425136751, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.516129493713379, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8534982204437256, "num_tokens": 202115273.0, "step": 5295 }, { "epoch": 0.6737056354153416, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.61282730102539, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8710353374481201, "num_tokens": 202145347.0, "step": 5296 }, { "epoch": 0.6738328456939321, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.541504859924316, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8656489253044128, "num_tokens": 202184621.0, "step": 5297 }, { "epoch": 0.6739600559725226, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.52212142944336, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8640937209129333, "num_tokens": 202230099.0, "step": 5298 }, { "epoch": 0.674087266251113, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.520442008972168, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8555659055709839, "num_tokens": 202267872.0, "step": 5299 }, { "epoch": 0.6742144765297036, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.540771484375, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.859824538230896, "num_tokens": 202305517.0, "step": 5300 }, { "epoch": 0.6743416868082941, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.603607177734375, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8469129800796509, "num_tokens": 202340671.0, "step": 5301 }, { "epoch": 0.6744688970868846, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.570155143737793, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8689807653427124, "num_tokens": 202379552.0, "step": 5302 }, { "epoch": 0.6745961073654752, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.53906536102295, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8558544516563416, "num_tokens": 202421404.0, "step": 5303 }, { "epoch": 0.6747233176440657, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.540002822875977, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8465602397918701, "num_tokens": 202463765.0, "step": 5304 }, { "epoch": 0.6748505279226561, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.552129745483398, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8562172055244446, "num_tokens": 202505872.0, "step": 5305 }, { "epoch": 0.6749777382012466, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.573164939880371, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8584938049316406, "num_tokens": 202544085.0, "step": 5306 }, { "epoch": 0.6751049484798372, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.51506519317627, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8567968606948853, "num_tokens": 202588237.0, "step": 5307 }, { "epoch": 0.6752321587584277, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.60252857208252, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8539749383926392, "num_tokens": 202628462.0, "step": 5308 }, { "epoch": 0.6753593690370182, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.583147048950195, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8628641963005066, "num_tokens": 202668149.0, "step": 5309 }, { "epoch": 0.6754865793156087, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.574129104614258, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8762864470481873, "num_tokens": 202708651.0, "step": 5310 }, { "epoch": 0.6756137895941993, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.667647361755371, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8653169870376587, "num_tokens": 202738613.0, "step": 5311 }, { "epoch": 0.6757409998727897, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.604451179504395, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8569731712341309, "num_tokens": 202779279.0, "step": 5312 }, { "epoch": 0.6758682101513802, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.606720924377441, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.86178058385849, "num_tokens": 202812095.0, "step": 5313 }, { "epoch": 0.6759954204299707, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.531731605529785, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8567129969596863, "num_tokens": 202855705.0, "step": 5314 }, { "epoch": 0.6761226307085613, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.56424331665039, "learning_rate": 1e-06, "loss": 0.4835, "mean_token_accuracy": 0.8439631462097168, "num_tokens": 202894180.0, "step": 5315 }, { "epoch": 0.6762498409871518, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.605411529541016, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8683645725250244, "num_tokens": 202925842.0, "step": 5316 }, { "epoch": 0.6763770512657423, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.610299110412598, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8657656908035278, "num_tokens": 202959103.0, "step": 5317 }, { "epoch": 0.6765042615443327, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.563129425048828, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8611556887626648, "num_tokens": 203000393.0, "step": 5318 }, { "epoch": 0.6766314718229233, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.587885856628418, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.861601710319519, "num_tokens": 203041613.0, "step": 5319 }, { "epoch": 0.6767586821015138, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.628265380859375, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8763771653175354, "num_tokens": 203073744.0, "step": 5320 }, { "epoch": 0.6768858923801043, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.550155639648438, "learning_rate": 1e-06, "loss": 0.4719, "mean_token_accuracy": 0.8459832072257996, "num_tokens": 203108350.0, "step": 5321 }, { "epoch": 0.6770131026586949, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.620224952697754, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8641840815544128, "num_tokens": 203139149.0, "step": 5322 }, { "epoch": 0.6771403129372854, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.561186790466309, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.844287633895874, "num_tokens": 203173953.0, "step": 5323 }, { "epoch": 0.6772675232158758, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.57889175415039, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8712229132652283, "num_tokens": 203208915.0, "step": 5324 }, { "epoch": 0.6773947334944663, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.568577766418457, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8487610816955566, "num_tokens": 203251906.0, "step": 5325 }, { "epoch": 0.6775219437730569, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.619573593139648, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8608278036117554, "num_tokens": 203289460.0, "step": 5326 }, { "epoch": 0.6776491540516474, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.56953239440918, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8691635131835938, "num_tokens": 203325647.0, "step": 5327 }, { "epoch": 0.6777763643302379, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.626216888427734, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8571362495422363, "num_tokens": 203364972.0, "step": 5328 }, { "epoch": 0.6779035746088284, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.543062210083008, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8610497713088989, "num_tokens": 203407778.0, "step": 5329 }, { "epoch": 0.6780307848874189, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.577483177185059, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.861859917640686, "num_tokens": 203443820.0, "step": 5330 }, { "epoch": 0.6781579951660094, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.595234870910645, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8665900230407715, "num_tokens": 203475742.0, "step": 5331 }, { "epoch": 0.6782852054445999, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.55305004119873, "learning_rate": 1e-06, "loss": 0.5119, "mean_token_accuracy": 0.8352353572845459, "num_tokens": 203512617.0, "step": 5332 }, { "epoch": 0.6784124157231904, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.658677101135254, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8451651334762573, "num_tokens": 203546778.0, "step": 5333 }, { "epoch": 0.678539626001781, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.581570625305176, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8629030585289001, "num_tokens": 203585222.0, "step": 5334 }, { "epoch": 0.6786668362803715, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.614153861999512, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8546804785728455, "num_tokens": 203627014.0, "step": 5335 }, { "epoch": 0.6787940465589619, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.56379508972168, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.865962028503418, "num_tokens": 203659189.0, "step": 5336 }, { "epoch": 0.6789212568375524, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.557573318481445, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8721539974212646, "num_tokens": 203695355.0, "step": 5337 }, { "epoch": 0.679048467116143, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.600990295410156, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8684377670288086, "num_tokens": 203733446.0, "step": 5338 }, { "epoch": 0.6791756773947335, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.593718528747559, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.844611406326294, "num_tokens": 203775387.0, "step": 5339 }, { "epoch": 0.679302887673324, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.570362091064453, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8563440442085266, "num_tokens": 203810713.0, "step": 5340 }, { "epoch": 0.6794300979519146, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.602021217346191, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8733903169631958, "num_tokens": 203847144.0, "step": 5341 }, { "epoch": 0.679557308230505, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.480003356933594, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8660892248153687, "num_tokens": 203889716.0, "step": 5342 }, { "epoch": 0.6796845185090955, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.637737274169922, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8473249673843384, "num_tokens": 203927692.0, "step": 5343 }, { "epoch": 0.679811728787686, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.507923126220703, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8479796051979065, "num_tokens": 203960180.0, "step": 5344 }, { "epoch": 0.6799389390662766, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.542437553405762, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8722708225250244, "num_tokens": 203997375.0, "step": 5345 }, { "epoch": 0.6800661493448671, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.58529281616211, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8574370741844177, "num_tokens": 204037588.0, "step": 5346 }, { "epoch": 0.6801933596234576, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.507098197937012, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.8534524440765381, "num_tokens": 204080993.0, "step": 5347 }, { "epoch": 0.680320569902048, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.523909568786621, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8635556697845459, "num_tokens": 204116863.0, "step": 5348 }, { "epoch": 0.6804477801806386, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.608038902282715, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8553785681724548, "num_tokens": 204156551.0, "step": 5349 }, { "epoch": 0.6805749904592291, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.514127731323242, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8544880151748657, "num_tokens": 204196210.0, "step": 5350 }, { "epoch": 0.6807022007378196, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.54942512512207, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8590008020401001, "num_tokens": 204234791.0, "step": 5351 }, { "epoch": 0.6808294110164101, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.578690528869629, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8647719621658325, "num_tokens": 204271631.0, "step": 5352 }, { "epoch": 0.6809566212950007, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.59274959564209, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8711555004119873, "num_tokens": 204311879.0, "step": 5353 }, { "epoch": 0.6810838315735911, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.496053695678711, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8641246557235718, "num_tokens": 204357123.0, "step": 5354 }, { "epoch": 0.6812110418521816, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.561863899230957, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8584128618240356, "num_tokens": 204397196.0, "step": 5355 }, { "epoch": 0.6813382521307721, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.581936836242676, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8518608212471008, "num_tokens": 204434654.0, "step": 5356 }, { "epoch": 0.6814654624093627, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.547379493713379, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8828282356262207, "num_tokens": 204466581.0, "step": 5357 }, { "epoch": 0.6815926726879532, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.561349868774414, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8738238215446472, "num_tokens": 204503890.0, "step": 5358 }, { "epoch": 0.6817198829665437, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.633415222167969, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8563781976699829, "num_tokens": 204537160.0, "step": 5359 }, { "epoch": 0.6818470932451343, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.546473503112793, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8716089725494385, "num_tokens": 204570077.0, "step": 5360 }, { "epoch": 0.6819743035237247, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.584176063537598, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8763428330421448, "num_tokens": 204603638.0, "step": 5361 }, { "epoch": 0.6821015138023152, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.52250862121582, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.857079267501831, "num_tokens": 204644584.0, "step": 5362 }, { "epoch": 0.6822287240809057, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.602856636047363, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.846979022026062, "num_tokens": 204678459.0, "step": 5363 }, { "epoch": 0.6823559343594963, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.543771743774414, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8585203886032104, "num_tokens": 204712923.0, "step": 5364 }, { "epoch": 0.6824831446380868, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.600117683410645, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8437459468841553, "num_tokens": 204751303.0, "step": 5365 }, { "epoch": 0.6826103549166773, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.58147144317627, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8611464500427246, "num_tokens": 204791685.0, "step": 5366 }, { "epoch": 0.6827375651952677, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.648295402526855, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8623223900794983, "num_tokens": 204827748.0, "step": 5367 }, { "epoch": 0.6828647754738583, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.604630470275879, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8773209452629089, "num_tokens": 204866568.0, "step": 5368 }, { "epoch": 0.6829919857524488, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.650187492370605, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8597002029418945, "num_tokens": 204908334.0, "step": 5369 }, { "epoch": 0.6831191960310393, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.53286361694336, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8641602993011475, "num_tokens": 204949929.0, "step": 5370 }, { "epoch": 0.6832464063096299, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.621767044067383, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8516251444816589, "num_tokens": 204991435.0, "step": 5371 }, { "epoch": 0.6833736165882204, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.611950874328613, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8671424388885498, "num_tokens": 205028434.0, "step": 5372 }, { "epoch": 0.6835008268668108, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.610162734985352, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8558158874511719, "num_tokens": 205064138.0, "step": 5373 }, { "epoch": 0.6836280371454013, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.657866477966309, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8553791046142578, "num_tokens": 205104058.0, "step": 5374 }, { "epoch": 0.6837552474239919, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.586592674255371, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8595513701438904, "num_tokens": 205142609.0, "step": 5375 }, { "epoch": 0.6838824577025824, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.612178802490234, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8717223405838013, "num_tokens": 205180407.0, "step": 5376 }, { "epoch": 0.6840096679811729, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.655106544494629, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8822466731071472, "num_tokens": 205217340.0, "step": 5377 }, { "epoch": 0.6841368782597634, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.608173370361328, "learning_rate": 1e-06, "loss": 0.4871, "mean_token_accuracy": 0.8496307730674744, "num_tokens": 205263671.0, "step": 5378 }, { "epoch": 0.6842640885383539, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.626734733581543, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8581846952438354, "num_tokens": 205302970.0, "step": 5379 }, { "epoch": 0.6843912988169444, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.636028289794922, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8650645613670349, "num_tokens": 205342247.0, "step": 5380 }, { "epoch": 0.6845185090955349, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.679208755493164, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8456078767776489, "num_tokens": 205374897.0, "step": 5381 }, { "epoch": 0.6846457193741254, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.627245903015137, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8577317595481873, "num_tokens": 205411645.0, "step": 5382 }, { "epoch": 0.684772929652716, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.587632179260254, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8610152006149292, "num_tokens": 205450972.0, "step": 5383 }, { "epoch": 0.6849001399313065, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.610372543334961, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8622000217437744, "num_tokens": 205494314.0, "step": 5384 }, { "epoch": 0.6850273502098969, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.626559257507324, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8480646014213562, "num_tokens": 205534696.0, "step": 5385 }, { "epoch": 0.6851545604884874, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.606139183044434, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8723801374435425, "num_tokens": 205575579.0, "step": 5386 }, { "epoch": 0.685281770767078, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.624201774597168, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8642919063568115, "num_tokens": 205609249.0, "step": 5387 }, { "epoch": 0.6854089810456685, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.657424926757812, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.857719898223877, "num_tokens": 205644135.0, "step": 5388 }, { "epoch": 0.685536191324259, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.58374309539795, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8548330068588257, "num_tokens": 205685312.0, "step": 5389 }, { "epoch": 0.6856634016028496, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.662027359008789, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8552234172821045, "num_tokens": 205720525.0, "step": 5390 }, { "epoch": 0.68579061188144, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.637861251831055, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.841019868850708, "num_tokens": 205763358.0, "step": 5391 }, { "epoch": 0.6859178221600305, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.644805908203125, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8634834885597229, "num_tokens": 205801413.0, "step": 5392 }, { "epoch": 0.686045032438621, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.71261215209961, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.862720787525177, "num_tokens": 205840442.0, "step": 5393 }, { "epoch": 0.6861722427172116, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.562569618225098, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8740758895874023, "num_tokens": 205877319.0, "step": 5394 }, { "epoch": 0.6862994529958021, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.674508094787598, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8592219352722168, "num_tokens": 205916186.0, "step": 5395 }, { "epoch": 0.6864266632743926, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.620278358459473, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8506684303283691, "num_tokens": 205956336.0, "step": 5396 }, { "epoch": 0.686553873552983, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.635538101196289, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8594155311584473, "num_tokens": 205989570.0, "step": 5397 }, { "epoch": 0.6866810838315736, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.650938034057617, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8535457849502563, "num_tokens": 206026139.0, "step": 5398 }, { "epoch": 0.6868082941101641, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.585752487182617, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8547139763832092, "num_tokens": 206070059.0, "step": 5399 }, { "epoch": 0.6869355043887546, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.610445022583008, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8638826608657837, "num_tokens": 206104645.0, "step": 5400 }, { "epoch": 0.6870627146673451, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.612651824951172, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8581224679946899, "num_tokens": 206141220.0, "step": 5401 }, { "epoch": 0.6871899249459357, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.620972633361816, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8582943081855774, "num_tokens": 206177917.0, "step": 5402 }, { "epoch": 0.6873171352245261, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.616134643554688, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8599385619163513, "num_tokens": 206213419.0, "step": 5403 }, { "epoch": 0.6874443455031166, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.638866424560547, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8447405099868774, "num_tokens": 206252421.0, "step": 5404 }, { "epoch": 0.6875715557817071, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.630108833312988, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8650164604187012, "num_tokens": 206287195.0, "step": 5405 }, { "epoch": 0.6876987660602977, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.645784378051758, "learning_rate": 1e-06, "loss": 0.5016, "mean_token_accuracy": 0.8409292697906494, "num_tokens": 206328339.0, "step": 5406 }, { "epoch": 0.6878259763388882, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.66140365600586, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8456611633300781, "num_tokens": 206369523.0, "step": 5407 }, { "epoch": 0.6879531866174787, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.614622116088867, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.879828691482544, "num_tokens": 206402864.0, "step": 5408 }, { "epoch": 0.6880803968960693, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.6876802444458, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8486536741256714, "num_tokens": 206439004.0, "step": 5409 }, { "epoch": 0.6882076071746597, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.63914966583252, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.862541675567627, "num_tokens": 206472320.0, "step": 5410 }, { "epoch": 0.6883348174532502, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.606708526611328, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8531409502029419, "num_tokens": 206518649.0, "step": 5411 }, { "epoch": 0.6884620277318407, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.712482452392578, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8747265338897705, "num_tokens": 206551318.0, "step": 5412 }, { "epoch": 0.6885892380104313, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.59586238861084, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8487836122512817, "num_tokens": 206588173.0, "step": 5413 }, { "epoch": 0.6887164482890218, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.61538314819336, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.858097493648529, "num_tokens": 206624013.0, "step": 5414 }, { "epoch": 0.6888436585676123, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.628961563110352, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8676626682281494, "num_tokens": 206668180.0, "step": 5415 }, { "epoch": 0.6889708688462027, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.700366020202637, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8557668924331665, "num_tokens": 206710619.0, "step": 5416 }, { "epoch": 0.6890980791247933, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.606097221374512, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8650301098823547, "num_tokens": 206750121.0, "step": 5417 }, { "epoch": 0.6892252894033838, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.700374603271484, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8713265657424927, "num_tokens": 206789538.0, "step": 5418 }, { "epoch": 0.6893524996819743, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.638062477111816, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8605201244354248, "num_tokens": 206827513.0, "step": 5419 }, { "epoch": 0.6894797099605648, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.64421272277832, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.867180585861206, "num_tokens": 206866563.0, "step": 5420 }, { "epoch": 0.6896069202391554, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.640596389770508, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8706990480422974, "num_tokens": 206906281.0, "step": 5421 }, { "epoch": 0.6897341305177458, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.616410255432129, "learning_rate": 1e-06, "loss": 0.4853, "mean_token_accuracy": 0.8418130874633789, "num_tokens": 206944055.0, "step": 5422 }, { "epoch": 0.6898613407963363, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.650262832641602, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8611147403717041, "num_tokens": 206984421.0, "step": 5423 }, { "epoch": 0.6899885510749268, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.639442443847656, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8494735956192017, "num_tokens": 207020319.0, "step": 5424 }, { "epoch": 0.6901157613535174, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.659031867980957, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8636425137519836, "num_tokens": 207061487.0, "step": 5425 }, { "epoch": 0.6902429716321079, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.689007759094238, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8620590567588806, "num_tokens": 207098343.0, "step": 5426 }, { "epoch": 0.6903701819106984, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.606468200683594, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8569505214691162, "num_tokens": 207140381.0, "step": 5427 }, { "epoch": 0.6904973921892888, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.639863967895508, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8737828731536865, "num_tokens": 207173923.0, "step": 5428 }, { "epoch": 0.6906246024678794, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.608329772949219, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8673019409179688, "num_tokens": 207214840.0, "step": 5429 }, { "epoch": 0.6907518127464699, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.69691276550293, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8597632646560669, "num_tokens": 207247850.0, "step": 5430 }, { "epoch": 0.6908790230250604, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.668920516967773, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8609629273414612, "num_tokens": 207286698.0, "step": 5431 }, { "epoch": 0.691006233303651, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.611629486083984, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8554477691650391, "num_tokens": 207322470.0, "step": 5432 }, { "epoch": 0.6911334435822415, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.675033569335938, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8467289209365845, "num_tokens": 207361517.0, "step": 5433 }, { "epoch": 0.6912606538608319, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.67385196685791, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8695259094238281, "num_tokens": 207394504.0, "step": 5434 }, { "epoch": 0.6913878641394224, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.795296669006348, "learning_rate": 1e-06, "loss": 0.4894, "mean_token_accuracy": 0.8423260450363159, "num_tokens": 207430148.0, "step": 5435 }, { "epoch": 0.691515074418013, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.626689910888672, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8580104112625122, "num_tokens": 207461094.0, "step": 5436 }, { "epoch": 0.6916422846966035, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.683836936950684, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8671362400054932, "num_tokens": 207500572.0, "step": 5437 }, { "epoch": 0.691769494975194, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.707444190979004, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8530082702636719, "num_tokens": 207534977.0, "step": 5438 }, { "epoch": 0.6918967052537845, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.661009788513184, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8660381436347961, "num_tokens": 207569809.0, "step": 5439 }, { "epoch": 0.692023915532375, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.632246017456055, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8619588017463684, "num_tokens": 207606785.0, "step": 5440 }, { "epoch": 0.6921511258109655, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.650470733642578, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8592777252197266, "num_tokens": 207648344.0, "step": 5441 }, { "epoch": 0.692278336089556, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.708183288574219, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8681824207305908, "num_tokens": 207680962.0, "step": 5442 }, { "epoch": 0.6924055463681466, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.661344528198242, "learning_rate": 1e-06, "loss": 0.3252, "mean_token_accuracy": 0.8930371999740601, "num_tokens": 207709162.0, "step": 5443 }, { "epoch": 0.6925327566467371, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.603577613830566, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8568432331085205, "num_tokens": 207746089.0, "step": 5444 }, { "epoch": 0.6926599669253276, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.698880195617676, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8721234798431396, "num_tokens": 207791235.0, "step": 5445 }, { "epoch": 0.692787177203918, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.708196640014648, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.852423906326294, "num_tokens": 207824527.0, "step": 5446 }, { "epoch": 0.6929143874825086, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.651504516601562, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8715699911117554, "num_tokens": 207863535.0, "step": 5447 }, { "epoch": 0.6930415977610991, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.697447776794434, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8553265333175659, "num_tokens": 207899694.0, "step": 5448 }, { "epoch": 0.6931688080396896, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.615124702453613, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.852490246295929, "num_tokens": 207931440.0, "step": 5449 }, { "epoch": 0.6932960183182801, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.752079963684082, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8598426580429077, "num_tokens": 207963758.0, "step": 5450 }, { "epoch": 0.6934232285968707, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.607987403869629, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8658448457717896, "num_tokens": 208002383.0, "step": 5451 }, { "epoch": 0.6935504388754611, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.69154167175293, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8649536371231079, "num_tokens": 208038723.0, "step": 5452 }, { "epoch": 0.6936776491540516, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.726792335510254, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8661901950836182, "num_tokens": 208071851.0, "step": 5453 }, { "epoch": 0.6938048594326421, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.685917854309082, "learning_rate": 1e-06, "loss": 0.4986, "mean_token_accuracy": 0.8423508405685425, "num_tokens": 208106916.0, "step": 5454 }, { "epoch": 0.6939320697112327, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.673942565917969, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8711970448493958, "num_tokens": 208146883.0, "step": 5455 }, { "epoch": 0.6940592799898232, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.724859237670898, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8632517457008362, "num_tokens": 208187364.0, "step": 5456 }, { "epoch": 0.6941864902684137, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.641257286071777, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8679383993148804, "num_tokens": 208226087.0, "step": 5457 }, { "epoch": 0.6943137005470043, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.643586158752441, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.858493447303772, "num_tokens": 208265193.0, "step": 5458 }, { "epoch": 0.6944409108255947, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.647869110107422, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8747055530548096, "num_tokens": 208305163.0, "step": 5459 }, { "epoch": 0.6945681211041852, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.67655086517334, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8501583337783813, "num_tokens": 208342229.0, "step": 5460 }, { "epoch": 0.6946953313827757, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.604640007019043, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8517295122146606, "num_tokens": 208384606.0, "step": 5461 }, { "epoch": 0.6948225416613663, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.661465644836426, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8794666528701782, "num_tokens": 208416777.0, "step": 5462 }, { "epoch": 0.6949497519399568, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.662690162658691, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8463893532752991, "num_tokens": 208456101.0, "step": 5463 }, { "epoch": 0.6950769622185473, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.607596397399902, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8716062307357788, "num_tokens": 208489201.0, "step": 5464 }, { "epoch": 0.6952041724971377, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.708812713623047, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8754156231880188, "num_tokens": 208525006.0, "step": 5465 }, { "epoch": 0.6953313827757283, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.625768661499023, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8580994606018066, "num_tokens": 208568997.0, "step": 5466 }, { "epoch": 0.6954585930543188, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.662055969238281, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8537581562995911, "num_tokens": 208605925.0, "step": 5467 }, { "epoch": 0.6955858033329093, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.694709777832031, "learning_rate": 1e-06, "loss": 0.5195, "mean_token_accuracy": 0.8362898826599121, "num_tokens": 208646290.0, "step": 5468 }, { "epoch": 0.6957130136114998, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.629071235656738, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8623372316360474, "num_tokens": 208686424.0, "step": 5469 }, { "epoch": 0.6958402238900904, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.679688453674316, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8624338507652283, "num_tokens": 208719643.0, "step": 5470 }, { "epoch": 0.6959674341686808, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.696980476379395, "learning_rate": 1e-06, "loss": 0.4915, "mean_token_accuracy": 0.845474123954773, "num_tokens": 208760804.0, "step": 5471 }, { "epoch": 0.6960946444472713, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.663533210754395, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.855642557144165, "num_tokens": 208803022.0, "step": 5472 }, { "epoch": 0.6962218547258618, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.733510971069336, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8613648414611816, "num_tokens": 208844205.0, "step": 5473 }, { "epoch": 0.6963490650044524, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.655729293823242, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8716628551483154, "num_tokens": 208886882.0, "step": 5474 }, { "epoch": 0.6964762752830429, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.729891777038574, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8581275939941406, "num_tokens": 208926649.0, "step": 5475 }, { "epoch": 0.6966034855616334, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.643938064575195, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8552566766738892, "num_tokens": 208963403.0, "step": 5476 }, { "epoch": 0.6967306958402238, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.680169105529785, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8531573414802551, "num_tokens": 209002475.0, "step": 5477 }, { "epoch": 0.6968579061188144, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.659806251525879, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8609091639518738, "num_tokens": 209044216.0, "step": 5478 }, { "epoch": 0.6969851163974049, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.646185874938965, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8507784605026245, "num_tokens": 209080774.0, "step": 5479 }, { "epoch": 0.6971123266759954, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.733928680419922, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8668620586395264, "num_tokens": 209117167.0, "step": 5480 }, { "epoch": 0.697239536954586, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.678462982177734, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8552141189575195, "num_tokens": 209148872.0, "step": 5481 }, { "epoch": 0.6973667472331765, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.629016876220703, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8627786636352539, "num_tokens": 209187912.0, "step": 5482 }, { "epoch": 0.6974939575117669, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.674134254455566, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8679232001304626, "num_tokens": 209227055.0, "step": 5483 }, { "epoch": 0.6976211677903574, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.685379981994629, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8743733167648315, "num_tokens": 209255343.0, "step": 5484 }, { "epoch": 0.697748378068948, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.711431503295898, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8538417816162109, "num_tokens": 209294621.0, "step": 5485 }, { "epoch": 0.6978755883475385, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.612208366394043, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8765807151794434, "num_tokens": 209336519.0, "step": 5486 }, { "epoch": 0.698002798626129, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.665264129638672, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8596032857894897, "num_tokens": 209372913.0, "step": 5487 }, { "epoch": 0.6981300089047195, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.601217269897461, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8530862331390381, "num_tokens": 209417768.0, "step": 5488 }, { "epoch": 0.69825721918331, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.6636323928833, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8639038801193237, "num_tokens": 209451221.0, "step": 5489 }, { "epoch": 0.6983844294619005, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.687542915344238, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8595936298370361, "num_tokens": 209488710.0, "step": 5490 }, { "epoch": 0.698511639740491, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.641337394714355, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8576117157936096, "num_tokens": 209525166.0, "step": 5491 }, { "epoch": 0.6986388500190815, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.681946754455566, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8636033535003662, "num_tokens": 209564804.0, "step": 5492 }, { "epoch": 0.6987660602976721, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.598105430603027, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8650314807891846, "num_tokens": 209607155.0, "step": 5493 }, { "epoch": 0.6988932705762626, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.62686824798584, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8501483201980591, "num_tokens": 209641226.0, "step": 5494 }, { "epoch": 0.699020480854853, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.654454231262207, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8594584465026855, "num_tokens": 209679180.0, "step": 5495 }, { "epoch": 0.6991476911334435, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.631123542785645, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8518636226654053, "num_tokens": 209716412.0, "step": 5496 }, { "epoch": 0.6992749014120341, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.733640670776367, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8517611026763916, "num_tokens": 209755697.0, "step": 5497 }, { "epoch": 0.6994021116906246, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.637533187866211, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8677936792373657, "num_tokens": 209793654.0, "step": 5498 }, { "epoch": 0.6995293219692151, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.604363441467285, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8594410419464111, "num_tokens": 209830703.0, "step": 5499 }, { "epoch": 0.6996565322478057, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.637984275817871, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.860623300075531, "num_tokens": 209868994.0, "step": 5500 }, { "epoch": 0.6997837425263961, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.619341850280762, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8701590299606323, "num_tokens": 209909605.0, "step": 5501 }, { "epoch": 0.6999109528049866, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.592787742614746, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8674226403236389, "num_tokens": 209945196.0, "step": 5502 }, { "epoch": 0.7000381630835771, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.631747245788574, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.86134934425354, "num_tokens": 209988585.0, "step": 5503 }, { "epoch": 0.7001653733621677, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.6397123336792, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.86081862449646, "num_tokens": 210032084.0, "step": 5504 }, { "epoch": 0.7002925836407582, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.637068748474121, "learning_rate": 1e-06, "loss": 0.4737, "mean_token_accuracy": 0.8502532839775085, "num_tokens": 210070639.0, "step": 5505 }, { "epoch": 0.7004197939193487, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.652908325195312, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8691622018814087, "num_tokens": 210113793.0, "step": 5506 }, { "epoch": 0.7005470041979391, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.690056800842285, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8534768223762512, "num_tokens": 210153517.0, "step": 5507 }, { "epoch": 0.7006742144765297, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.677967071533203, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8564431071281433, "num_tokens": 210186818.0, "step": 5508 }, { "epoch": 0.7008014247551202, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.651969909667969, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8531917333602905, "num_tokens": 210224283.0, "step": 5509 }, { "epoch": 0.7009286350337107, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.684229850769043, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8551976680755615, "num_tokens": 210263439.0, "step": 5510 }, { "epoch": 0.7010558453123013, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.671162605285645, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8627454042434692, "num_tokens": 210305887.0, "step": 5511 }, { "epoch": 0.7011830555908918, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.646660804748535, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8598495125770569, "num_tokens": 210340118.0, "step": 5512 }, { "epoch": 0.7013102658694823, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.661396026611328, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8616509437561035, "num_tokens": 210376098.0, "step": 5513 }, { "epoch": 0.7014374761480727, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.67188549041748, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.864057183265686, "num_tokens": 210418801.0, "step": 5514 }, { "epoch": 0.7015646864266633, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.669137954711914, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8503648638725281, "num_tokens": 210452392.0, "step": 5515 }, { "epoch": 0.7016918967052538, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.668262481689453, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.868071436882019, "num_tokens": 210489697.0, "step": 5516 }, { "epoch": 0.7018191069838443, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.65181827545166, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.845289945602417, "num_tokens": 210530212.0, "step": 5517 }, { "epoch": 0.7019463172624348, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.671311378479004, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.866836428642273, "num_tokens": 210564974.0, "step": 5518 }, { "epoch": 0.7020735275410254, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.635519981384277, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8621023893356323, "num_tokens": 210601272.0, "step": 5519 }, { "epoch": 0.7022007378196158, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.628174781799316, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8634556531906128, "num_tokens": 210636908.0, "step": 5520 }, { "epoch": 0.7023279480982063, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.68218994140625, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.857279896736145, "num_tokens": 210672629.0, "step": 5521 }, { "epoch": 0.7024551583767968, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.665911674499512, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.875302255153656, "num_tokens": 210703531.0, "step": 5522 }, { "epoch": 0.7025823686553874, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.687054634094238, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8734607100486755, "num_tokens": 210742946.0, "step": 5523 }, { "epoch": 0.7027095789339779, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.597928047180176, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8698675632476807, "num_tokens": 210778345.0, "step": 5524 }, { "epoch": 0.7028367892125684, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.704414367675781, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8669177293777466, "num_tokens": 210821544.0, "step": 5525 }, { "epoch": 0.7029639994911588, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.645966529846191, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8648239970207214, "num_tokens": 210858060.0, "step": 5526 }, { "epoch": 0.7030912097697494, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.72240924835205, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8556133508682251, "num_tokens": 210900365.0, "step": 5527 }, { "epoch": 0.7032184200483399, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.639774322509766, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8560488224029541, "num_tokens": 210941118.0, "step": 5528 }, { "epoch": 0.7033456303269304, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.660239219665527, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8681851029396057, "num_tokens": 210979865.0, "step": 5529 }, { "epoch": 0.703472840605521, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.637117385864258, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8614561557769775, "num_tokens": 211016385.0, "step": 5530 }, { "epoch": 0.7036000508841115, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.64705753326416, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.8494864702224731, "num_tokens": 211059071.0, "step": 5531 }, { "epoch": 0.7037272611627019, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.63524055480957, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8640813827514648, "num_tokens": 211094189.0, "step": 5532 }, { "epoch": 0.7038544714412924, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.679353713989258, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8756206631660461, "num_tokens": 211130524.0, "step": 5533 }, { "epoch": 0.703981681719883, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.629532814025879, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8546865582466125, "num_tokens": 211172335.0, "step": 5534 }, { "epoch": 0.7041088919984735, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.666409492492676, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8635867834091187, "num_tokens": 211205406.0, "step": 5535 }, { "epoch": 0.704236102277064, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.675724983215332, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8568530678749084, "num_tokens": 211246443.0, "step": 5536 }, { "epoch": 0.7043633125556545, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.650001525878906, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8754392862319946, "num_tokens": 211286707.0, "step": 5537 }, { "epoch": 0.704490522834245, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.700733184814453, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8720197081565857, "num_tokens": 211319448.0, "step": 5538 }, { "epoch": 0.7046177331128355, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.63072395324707, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8559277057647705, "num_tokens": 211363747.0, "step": 5539 }, { "epoch": 0.704744943391426, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.674242973327637, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8680078983306885, "num_tokens": 211408080.0, "step": 5540 }, { "epoch": 0.7048721536700165, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.669647216796875, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8509873747825623, "num_tokens": 211445656.0, "step": 5541 }, { "epoch": 0.7049993639486071, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.663211822509766, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8708606958389282, "num_tokens": 211484342.0, "step": 5542 }, { "epoch": 0.7051265742271976, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.682733535766602, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8568210601806641, "num_tokens": 211518535.0, "step": 5543 }, { "epoch": 0.705253784505788, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.661995887756348, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8554089069366455, "num_tokens": 211557236.0, "step": 5544 }, { "epoch": 0.7053809947843785, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.672961235046387, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8724097013473511, "num_tokens": 211599216.0, "step": 5545 }, { "epoch": 0.7055082050629691, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.666231155395508, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8434479236602783, "num_tokens": 211638930.0, "step": 5546 }, { "epoch": 0.7056354153415596, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.69472599029541, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8538697361946106, "num_tokens": 211680765.0, "step": 5547 }, { "epoch": 0.7057626256201501, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.652996063232422, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8745318055152893, "num_tokens": 211719037.0, "step": 5548 }, { "epoch": 0.7058898358987407, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.678070068359375, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8560194969177246, "num_tokens": 211759655.0, "step": 5549 }, { "epoch": 0.7060170461773311, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.715226173400879, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.8433430790901184, "num_tokens": 211800405.0, "step": 5550 }, { "epoch": 0.7061442564559216, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.6265230178833, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8434975147247314, "num_tokens": 211846057.0, "step": 5551 }, { "epoch": 0.7062714667345121, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.685688972473145, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8582204580307007, "num_tokens": 211885944.0, "step": 5552 }, { "epoch": 0.7063986770131027, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.655816078186035, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8654637336730957, "num_tokens": 211928727.0, "step": 5553 }, { "epoch": 0.7065258872916932, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.700175285339355, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8517273664474487, "num_tokens": 211967593.0, "step": 5554 }, { "epoch": 0.7066530975702837, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.650758743286133, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8559569120407104, "num_tokens": 212001023.0, "step": 5555 }, { "epoch": 0.7067803078488741, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.653504371643066, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8525781631469727, "num_tokens": 212045611.0, "step": 5556 }, { "epoch": 0.7069075181274647, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.6426420211792, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8563960790634155, "num_tokens": 212087167.0, "step": 5557 }, { "epoch": 0.7070347284060552, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.642923355102539, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8620256781578064, "num_tokens": 212124254.0, "step": 5558 }, { "epoch": 0.7071619386846457, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.69420337677002, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8555739521980286, "num_tokens": 212162706.0, "step": 5559 }, { "epoch": 0.7072891489632362, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.639888763427734, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.861107587814331, "num_tokens": 212197359.0, "step": 5560 }, { "epoch": 0.7074163592418268, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.664739608764648, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8674763441085815, "num_tokens": 212237300.0, "step": 5561 }, { "epoch": 0.7075435695204173, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.639174461364746, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8413287401199341, "num_tokens": 212283472.0, "step": 5562 }, { "epoch": 0.7076707797990077, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.674627304077148, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8547554612159729, "num_tokens": 212321433.0, "step": 5563 }, { "epoch": 0.7077979900775982, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.648658752441406, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8587528467178345, "num_tokens": 212362234.0, "step": 5564 }, { "epoch": 0.7079252003561888, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.696762084960938, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8505429029464722, "num_tokens": 212399259.0, "step": 5565 }, { "epoch": 0.7080524106347793, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.659723281860352, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8608337640762329, "num_tokens": 212438489.0, "step": 5566 }, { "epoch": 0.7081796209133698, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.66972541809082, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8524674773216248, "num_tokens": 212482283.0, "step": 5567 }, { "epoch": 0.7083068311919604, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.680768013000488, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.855300784111023, "num_tokens": 212522749.0, "step": 5568 }, { "epoch": 0.7084340414705508, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.659135818481445, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.851719081401825, "num_tokens": 212559937.0, "step": 5569 }, { "epoch": 0.7085612517491413, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.679512023925781, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8657758235931396, "num_tokens": 212595578.0, "step": 5570 }, { "epoch": 0.7086884620277318, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.619266510009766, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8718372583389282, "num_tokens": 212634085.0, "step": 5571 }, { "epoch": 0.7088156723063224, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.70544147491455, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8496582508087158, "num_tokens": 212673557.0, "step": 5572 }, { "epoch": 0.7089428825849129, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.627997398376465, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8699169158935547, "num_tokens": 212711142.0, "step": 5573 }, { "epoch": 0.7090700928635034, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.63785457611084, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8608523011207581, "num_tokens": 212745570.0, "step": 5574 }, { "epoch": 0.7091973031420938, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.613845825195312, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8635361194610596, "num_tokens": 212783048.0, "step": 5575 }, { "epoch": 0.7093245134206844, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.60300064086914, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8775581121444702, "num_tokens": 212821231.0, "step": 5576 }, { "epoch": 0.7094517236992749, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.61528491973877, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8588900566101074, "num_tokens": 212861715.0, "step": 5577 }, { "epoch": 0.7095789339778654, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.643860816955566, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8499190807342529, "num_tokens": 212900655.0, "step": 5578 }, { "epoch": 0.709706144256456, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.59257698059082, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8514800071716309, "num_tokens": 212939308.0, "step": 5579 }, { "epoch": 0.7098333545350465, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.681519508361816, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8640279769897461, "num_tokens": 212981174.0, "step": 5580 }, { "epoch": 0.7099605648136369, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.610010147094727, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.870144248008728, "num_tokens": 213016401.0, "step": 5581 }, { "epoch": 0.7100877750922274, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.70143985748291, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8483724594116211, "num_tokens": 213058480.0, "step": 5582 }, { "epoch": 0.710214985370818, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.676995277404785, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8729710578918457, "num_tokens": 213093527.0, "step": 5583 }, { "epoch": 0.7103421956494085, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.691963195800781, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8577190637588501, "num_tokens": 213126778.0, "step": 5584 }, { "epoch": 0.710469405927999, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.616914749145508, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8632525205612183, "num_tokens": 213170863.0, "step": 5585 }, { "epoch": 0.7105966162065895, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.637675285339355, "learning_rate": 1e-06, "loss": 0.4999, "mean_token_accuracy": 0.8364259004592896, "num_tokens": 213211336.0, "step": 5586 }, { "epoch": 0.71072382648518, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.616827011108398, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8681697845458984, "num_tokens": 213245262.0, "step": 5587 }, { "epoch": 0.7108510367637705, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.715009689331055, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8519410490989685, "num_tokens": 213282150.0, "step": 5588 }, { "epoch": 0.710978247042361, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.619111061096191, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.865410327911377, "num_tokens": 213325573.0, "step": 5589 }, { "epoch": 0.7111054573209515, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.65246295928955, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8681743144989014, "num_tokens": 213363738.0, "step": 5590 }, { "epoch": 0.7112326675995421, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.642792701721191, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8656910061836243, "num_tokens": 213404796.0, "step": 5591 }, { "epoch": 0.7113598778781326, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.616846084594727, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8526232838630676, "num_tokens": 213436711.0, "step": 5592 }, { "epoch": 0.711487088156723, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.690474510192871, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8626652956008911, "num_tokens": 213473458.0, "step": 5593 }, { "epoch": 0.7116142984353135, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.671503067016602, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.870163083076477, "num_tokens": 213507355.0, "step": 5594 }, { "epoch": 0.7117415087139041, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.661093711853027, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8436216115951538, "num_tokens": 213543771.0, "step": 5595 }, { "epoch": 0.7118687189924946, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.674057006835938, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8588072061538696, "num_tokens": 213578317.0, "step": 5596 }, { "epoch": 0.7119959292710851, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.608746528625488, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8565763235092163, "num_tokens": 213615795.0, "step": 5597 }, { "epoch": 0.7121231395496757, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.723633766174316, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8695391416549683, "num_tokens": 213655452.0, "step": 5598 }, { "epoch": 0.7122503498282661, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.722715377807617, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8518283367156982, "num_tokens": 213688079.0, "step": 5599 }, { "epoch": 0.7123775601068566, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.625317573547363, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.876494288444519, "num_tokens": 213725310.0, "step": 5600 }, { "epoch": 0.7125047703854471, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.732797622680664, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8543045520782471, "num_tokens": 213757950.0, "step": 5601 }, { "epoch": 0.7126319806640377, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.699698448181152, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.8484352827072144, "num_tokens": 213794146.0, "step": 5602 }, { "epoch": 0.7127591909426282, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.667977333068848, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8643409013748169, "num_tokens": 213834922.0, "step": 5603 }, { "epoch": 0.7128864012212187, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.683589935302734, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.878889799118042, "num_tokens": 213879651.0, "step": 5604 }, { "epoch": 0.7130136114998091, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.696252822875977, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.859192967414856, "num_tokens": 213921833.0, "step": 5605 }, { "epoch": 0.7131408217783997, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.663236618041992, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.864729642868042, "num_tokens": 213962909.0, "step": 5606 }, { "epoch": 0.7132680320569902, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.713932991027832, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8625738620758057, "num_tokens": 214002030.0, "step": 5607 }, { "epoch": 0.7133952423355807, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.663469314575195, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8536117076873779, "num_tokens": 214037094.0, "step": 5608 }, { "epoch": 0.7135224526141712, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.706392288208008, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8640976548194885, "num_tokens": 214076649.0, "step": 5609 }, { "epoch": 0.7136496628927618, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.637104034423828, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8567127585411072, "num_tokens": 214114194.0, "step": 5610 }, { "epoch": 0.7137768731713523, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.702265739440918, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.86240553855896, "num_tokens": 214149652.0, "step": 5611 }, { "epoch": 0.7139040834499427, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.636772155761719, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8700571656227112, "num_tokens": 214189068.0, "step": 5612 }, { "epoch": 0.7140312937285332, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.695528030395508, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8482224941253662, "num_tokens": 214220251.0, "step": 5613 }, { "epoch": 0.7141585040071238, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.702669143676758, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8581243753433228, "num_tokens": 214256705.0, "step": 5614 }, { "epoch": 0.7142857142857143, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.688984870910645, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8744982481002808, "num_tokens": 214292563.0, "step": 5615 }, { "epoch": 0.7144129245643048, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.669449806213379, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8548436164855957, "num_tokens": 214333843.0, "step": 5616 }, { "epoch": 0.7145401348428954, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.637688636779785, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8717492818832397, "num_tokens": 214372272.0, "step": 5617 }, { "epoch": 0.7146673451214858, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.62960147857666, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8671621680259705, "num_tokens": 214416364.0, "step": 5618 }, { "epoch": 0.7147945554000763, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.644503593444824, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8583650588989258, "num_tokens": 214453102.0, "step": 5619 }, { "epoch": 0.7149217656786668, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.673934936523438, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8644476532936096, "num_tokens": 214492374.0, "step": 5620 }, { "epoch": 0.7150489759572574, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.686327934265137, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8612781763076782, "num_tokens": 214530918.0, "step": 5621 }, { "epoch": 0.7151761862358479, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.632120132446289, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8771398067474365, "num_tokens": 214570227.0, "step": 5622 }, { "epoch": 0.7153033965144384, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.669157028198242, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8634737133979797, "num_tokens": 214609519.0, "step": 5623 }, { "epoch": 0.7154306067930288, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.692766189575195, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8573862910270691, "num_tokens": 214643564.0, "step": 5624 }, { "epoch": 0.7155578170716194, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.685162544250488, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8690168857574463, "num_tokens": 214677632.0, "step": 5625 }, { "epoch": 0.7156850273502099, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.66077995300293, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8630690574645996, "num_tokens": 214710363.0, "step": 5626 }, { "epoch": 0.7158122376288004, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.680887222290039, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8682553172111511, "num_tokens": 214746982.0, "step": 5627 }, { "epoch": 0.715939447907391, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.644423484802246, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8518489003181458, "num_tokens": 214786447.0, "step": 5628 }, { "epoch": 0.7160666581859815, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.742330551147461, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8691184520721436, "num_tokens": 214824627.0, "step": 5629 }, { "epoch": 0.7161938684645719, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.7091646194458, "learning_rate": 1e-06, "loss": 0.4973, "mean_token_accuracy": 0.842624843120575, "num_tokens": 214863051.0, "step": 5630 }, { "epoch": 0.7163210787431624, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.730972290039062, "learning_rate": 1e-06, "loss": 0.493, "mean_token_accuracy": 0.8479834198951721, "num_tokens": 214900524.0, "step": 5631 }, { "epoch": 0.716448289021753, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.663228034973145, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8666936159133911, "num_tokens": 214943949.0, "step": 5632 }, { "epoch": 0.7165754993003435, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.688528060913086, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.865942120552063, "num_tokens": 214979126.0, "step": 5633 }, { "epoch": 0.716702709578934, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.69786262512207, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8607705235481262, "num_tokens": 215017740.0, "step": 5634 }, { "epoch": 0.7168299198575245, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.624747276306152, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8697303533554077, "num_tokens": 215054472.0, "step": 5635 }, { "epoch": 0.716957130136115, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.69838809967041, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8643673658370972, "num_tokens": 215086894.0, "step": 5636 }, { "epoch": 0.7170843404147055, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.671440124511719, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8614723682403564, "num_tokens": 215128157.0, "step": 5637 }, { "epoch": 0.717211550693296, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.720455169677734, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8684197664260864, "num_tokens": 215161008.0, "step": 5638 }, { "epoch": 0.7173387609718865, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.693306922912598, "learning_rate": 1e-06, "loss": 0.5063, "mean_token_accuracy": 0.8448894023895264, "num_tokens": 215197157.0, "step": 5639 }, { "epoch": 0.7174659712504771, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.685236930847168, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8692588806152344, "num_tokens": 215240270.0, "step": 5640 }, { "epoch": 0.7175931815290676, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.685023307800293, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8686673641204834, "num_tokens": 215279828.0, "step": 5641 }, { "epoch": 0.717720391807658, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.924488067626953, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.855701744556427, "num_tokens": 215315271.0, "step": 5642 }, { "epoch": 0.7178476020862485, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.661874771118164, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8630070090293884, "num_tokens": 215352621.0, "step": 5643 }, { "epoch": 0.7179748123648391, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.646717071533203, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8637945652008057, "num_tokens": 215390231.0, "step": 5644 }, { "epoch": 0.7181020226434296, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.703975677490234, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8756667375564575, "num_tokens": 215426812.0, "step": 5645 }, { "epoch": 0.7182292329220201, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.686857223510742, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.8506482839584351, "num_tokens": 215467551.0, "step": 5646 }, { "epoch": 0.7183564432006107, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.705392837524414, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8700087666511536, "num_tokens": 215508559.0, "step": 5647 }, { "epoch": 0.7184836534792011, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.659404754638672, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8547311425209045, "num_tokens": 215548709.0, "step": 5648 }, { "epoch": 0.7186108637577916, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.704361915588379, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8623032569885254, "num_tokens": 215593844.0, "step": 5649 }, { "epoch": 0.7187380740363821, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.648882865905762, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8669048547744751, "num_tokens": 215633853.0, "step": 5650 }, { "epoch": 0.7188652843149727, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.7047700881958, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8608890771865845, "num_tokens": 215667272.0, "step": 5651 }, { "epoch": 0.7189924945935632, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.700565338134766, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.844744861125946, "num_tokens": 215705889.0, "step": 5652 }, { "epoch": 0.7191197048721537, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.674994468688965, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.870890200138092, "num_tokens": 215749143.0, "step": 5653 }, { "epoch": 0.7192469151507441, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.692841529846191, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8560678958892822, "num_tokens": 215787583.0, "step": 5654 }, { "epoch": 0.7193741254293347, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.753605842590332, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8613560199737549, "num_tokens": 215825552.0, "step": 5655 }, { "epoch": 0.7195013357079252, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.70439624786377, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8701135516166687, "num_tokens": 215858310.0, "step": 5656 }, { "epoch": 0.7196285459865157, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.744036674499512, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.850340723991394, "num_tokens": 215898403.0, "step": 5657 }, { "epoch": 0.7197557562651062, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.72756290435791, "learning_rate": 1e-06, "loss": 0.5232, "mean_token_accuracy": 0.8384830355644226, "num_tokens": 215935930.0, "step": 5658 }, { "epoch": 0.7198829665436968, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.708717346191406, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.858308732509613, "num_tokens": 215971509.0, "step": 5659 }, { "epoch": 0.7200101768222873, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.718794822692871, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8682956695556641, "num_tokens": 216008696.0, "step": 5660 }, { "epoch": 0.7201373871008777, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.730340003967285, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.861180305480957, "num_tokens": 216045861.0, "step": 5661 }, { "epoch": 0.7202645973794682, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.740297317504883, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8715376853942871, "num_tokens": 216081307.0, "step": 5662 }, { "epoch": 0.7203918076580588, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.6751708984375, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8699831366539001, "num_tokens": 216118506.0, "step": 5663 }, { "epoch": 0.7205190179366493, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.738218307495117, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8546714186668396, "num_tokens": 216159000.0, "step": 5664 }, { "epoch": 0.7206462282152398, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.648454666137695, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8544467687606812, "num_tokens": 216204712.0, "step": 5665 }, { "epoch": 0.7207734384938304, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.750883102416992, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8544985055923462, "num_tokens": 216235001.0, "step": 5666 }, { "epoch": 0.7209006487724208, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.766556739807129, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8552539944648743, "num_tokens": 216271383.0, "step": 5667 }, { "epoch": 0.7210278590510113, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.673734664916992, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8626459836959839, "num_tokens": 216307923.0, "step": 5668 }, { "epoch": 0.7211550693296018, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.672386169433594, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8808526992797852, "num_tokens": 216345147.0, "step": 5669 }, { "epoch": 0.7212822796081924, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.679121017456055, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8548203706741333, "num_tokens": 216386991.0, "step": 5670 }, { "epoch": 0.7214094898867829, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.763635635375977, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8645340204238892, "num_tokens": 216425534.0, "step": 5671 }, { "epoch": 0.7215367001653734, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.711599349975586, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.861092209815979, "num_tokens": 216466686.0, "step": 5672 }, { "epoch": 0.7216639104439638, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.732396125793457, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8733870983123779, "num_tokens": 216505245.0, "step": 5673 }, { "epoch": 0.7217911207225544, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.771458625793457, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8581265211105347, "num_tokens": 216538024.0, "step": 5674 }, { "epoch": 0.7219183310011449, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.732921600341797, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8604036569595337, "num_tokens": 216573722.0, "step": 5675 }, { "epoch": 0.7220455412797354, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.615350723266602, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8673782348632812, "num_tokens": 216610362.0, "step": 5676 }, { "epoch": 0.7221727515583259, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.773159980773926, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8537672162055969, "num_tokens": 216648732.0, "step": 5677 }, { "epoch": 0.7222999618369165, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.728707313537598, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8694567084312439, "num_tokens": 216680324.0, "step": 5678 }, { "epoch": 0.7224271721155069, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.700819969177246, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8648084402084351, "num_tokens": 216712773.0, "step": 5679 }, { "epoch": 0.7225543823940974, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.686976432800293, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8541795015335083, "num_tokens": 216752538.0, "step": 5680 }, { "epoch": 0.722681592672688, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.755756378173828, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8643306493759155, "num_tokens": 216791795.0, "step": 5681 }, { "epoch": 0.7228088029512785, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.717345237731934, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8579739928245544, "num_tokens": 216831729.0, "step": 5682 }, { "epoch": 0.722936013229869, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.728462219238281, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8618696928024292, "num_tokens": 216869494.0, "step": 5683 }, { "epoch": 0.7230632235084595, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.802979469299316, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.8434839248657227, "num_tokens": 216906672.0, "step": 5684 }, { "epoch": 0.72319043378705, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.732134819030762, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8686472177505493, "num_tokens": 216941587.0, "step": 5685 }, { "epoch": 0.7233176440656405, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.803701400756836, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8543473482131958, "num_tokens": 216981812.0, "step": 5686 }, { "epoch": 0.723444854344231, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.681182861328125, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8705778121948242, "num_tokens": 217024137.0, "step": 5687 }, { "epoch": 0.7235720646228215, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.731888771057129, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8736138343811035, "num_tokens": 217063753.0, "step": 5688 }, { "epoch": 0.7236992749014121, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.761240005493164, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8606086373329163, "num_tokens": 217105611.0, "step": 5689 }, { "epoch": 0.7238264851800026, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.716700553894043, "learning_rate": 1e-06, "loss": 0.4857, "mean_token_accuracy": 0.8418854475021362, "num_tokens": 217145989.0, "step": 5690 }, { "epoch": 0.723953695458593, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.73190975189209, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8793673515319824, "num_tokens": 217183726.0, "step": 5691 }, { "epoch": 0.7240809057371835, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.702975273132324, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8594903945922852, "num_tokens": 217226644.0, "step": 5692 }, { "epoch": 0.7242081160157741, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.728781700134277, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8472083806991577, "num_tokens": 217263327.0, "step": 5693 }, { "epoch": 0.7243353262943646, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.721390724182129, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8659945726394653, "num_tokens": 217300595.0, "step": 5694 }, { "epoch": 0.7244625365729551, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.782163619995117, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8477289080619812, "num_tokens": 217336385.0, "step": 5695 }, { "epoch": 0.7245897468515456, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.653604507446289, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8678311705589294, "num_tokens": 217374309.0, "step": 5696 }, { "epoch": 0.7247169571301361, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.663758277893066, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8530482053756714, "num_tokens": 217408386.0, "step": 5697 }, { "epoch": 0.7248441674087266, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.720443725585938, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.848671019077301, "num_tokens": 217442855.0, "step": 5698 }, { "epoch": 0.7249713776873171, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.66147232055664, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.86617112159729, "num_tokens": 217476892.0, "step": 5699 }, { "epoch": 0.7250985879659076, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.697617530822754, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8599404692649841, "num_tokens": 217513569.0, "step": 5700 }, { "epoch": 0.7252257982444982, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.675841331481934, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8565011024475098, "num_tokens": 217545585.0, "step": 5701 }, { "epoch": 0.7253530085230887, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.730648040771484, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8637206554412842, "num_tokens": 217583156.0, "step": 5702 }, { "epoch": 0.7254802188016791, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.705074310302734, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8620554208755493, "num_tokens": 217619630.0, "step": 5703 }, { "epoch": 0.7256074290802697, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.680846214294434, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8726475238800049, "num_tokens": 217653616.0, "step": 5704 }, { "epoch": 0.7257346393588602, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.69603443145752, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.857407808303833, "num_tokens": 217692758.0, "step": 5705 }, { "epoch": 0.7258618496374507, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.725436210632324, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8466159105300903, "num_tokens": 217734175.0, "step": 5706 }, { "epoch": 0.7259890599160412, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.73481559753418, "learning_rate": 1e-06, "loss": 0.4946, "mean_token_accuracy": 0.8390610218048096, "num_tokens": 217773372.0, "step": 5707 }, { "epoch": 0.7261162701946318, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.715341567993164, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8710139393806458, "num_tokens": 217809174.0, "step": 5708 }, { "epoch": 0.7262434804732223, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.691645622253418, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8545260429382324, "num_tokens": 217845882.0, "step": 5709 }, { "epoch": 0.7263706907518127, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.743329048156738, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8566211462020874, "num_tokens": 217885798.0, "step": 5710 }, { "epoch": 0.7264979010304032, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.758573532104492, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8583486676216125, "num_tokens": 217924967.0, "step": 5711 }, { "epoch": 0.7266251113089938, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.698235511779785, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8732910752296448, "num_tokens": 217967381.0, "step": 5712 }, { "epoch": 0.7267523215875843, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.745660781860352, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8674523830413818, "num_tokens": 218001097.0, "step": 5713 }, { "epoch": 0.7268795318661748, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.737858772277832, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8637537360191345, "num_tokens": 218040705.0, "step": 5714 }, { "epoch": 0.7270067421447653, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.74749755859375, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.861663281917572, "num_tokens": 218077549.0, "step": 5715 }, { "epoch": 0.7271339524233558, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.64371395111084, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8600285649299622, "num_tokens": 218116251.0, "step": 5716 }, { "epoch": 0.7272611627019463, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.726273536682129, "learning_rate": 1e-06, "loss": 0.5087, "mean_token_accuracy": 0.8354028463363647, "num_tokens": 218155798.0, "step": 5717 }, { "epoch": 0.7273883729805368, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.712264060974121, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8678663969039917, "num_tokens": 218191277.0, "step": 5718 }, { "epoch": 0.7275155832591274, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.7276029586792, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8669455051422119, "num_tokens": 218233165.0, "step": 5719 }, { "epoch": 0.7276427935377179, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.710779190063477, "learning_rate": 1e-06, "loss": 0.4954, "mean_token_accuracy": 0.8407394886016846, "num_tokens": 218271263.0, "step": 5720 }, { "epoch": 0.7277700038163084, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.759017944335938, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8634004592895508, "num_tokens": 218309081.0, "step": 5721 }, { "epoch": 0.7278972140948988, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.700410842895508, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8571358323097229, "num_tokens": 218348314.0, "step": 5722 }, { "epoch": 0.7280244243734894, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.726822853088379, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8642160296440125, "num_tokens": 218387556.0, "step": 5723 }, { "epoch": 0.7281516346520799, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.724238395690918, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.853793203830719, "num_tokens": 218423130.0, "step": 5724 }, { "epoch": 0.7282788449306704, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.797131538391113, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8434869050979614, "num_tokens": 218464079.0, "step": 5725 }, { "epoch": 0.7284060552092609, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.677521705627441, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8695086240768433, "num_tokens": 218501307.0, "step": 5726 }, { "epoch": 0.7285332654878515, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.79560661315918, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8683623671531677, "num_tokens": 218537710.0, "step": 5727 }, { "epoch": 0.7286604757664419, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.70159912109375, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.857280433177948, "num_tokens": 218576026.0, "step": 5728 }, { "epoch": 0.7287876860450324, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.721235275268555, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8574365377426147, "num_tokens": 218612491.0, "step": 5729 }, { "epoch": 0.7289148963236229, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.759517669677734, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8699321746826172, "num_tokens": 218642541.0, "step": 5730 }, { "epoch": 0.7290421066022135, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.733275413513184, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8476665019989014, "num_tokens": 218678514.0, "step": 5731 }, { "epoch": 0.729169316880804, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.760858535766602, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.866223931312561, "num_tokens": 218715415.0, "step": 5732 }, { "epoch": 0.7292965271593945, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.661194801330566, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8691518902778625, "num_tokens": 218754422.0, "step": 5733 }, { "epoch": 0.7294237374379849, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.79125690460205, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8525164127349854, "num_tokens": 218787141.0, "step": 5734 }, { "epoch": 0.7295509477165755, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.697113990783691, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8765362501144409, "num_tokens": 218824764.0, "step": 5735 }, { "epoch": 0.729678157995166, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.676084518432617, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8591426610946655, "num_tokens": 218866142.0, "step": 5736 }, { "epoch": 0.7298053682737565, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.71965503692627, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8568301200866699, "num_tokens": 218909208.0, "step": 5737 }, { "epoch": 0.7299325785523471, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.771897315979004, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8688114881515503, "num_tokens": 218942469.0, "step": 5738 }, { "epoch": 0.7300597888309376, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.780220985412598, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8540078401565552, "num_tokens": 218982934.0, "step": 5739 }, { "epoch": 0.730186999109528, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.719953536987305, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8581945896148682, "num_tokens": 219027055.0, "step": 5740 }, { "epoch": 0.7303142093881185, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.746288299560547, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8611189723014832, "num_tokens": 219069639.0, "step": 5741 }, { "epoch": 0.7304414196667091, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.737374305725098, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8551443815231323, "num_tokens": 219111774.0, "step": 5742 }, { "epoch": 0.7305686299452996, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.716765403747559, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8477870225906372, "num_tokens": 219156962.0, "step": 5743 }, { "epoch": 0.7306958402238901, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.703590393066406, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8703461289405823, "num_tokens": 219193352.0, "step": 5744 }, { "epoch": 0.7308230505024806, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.831099510192871, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8463740348815918, "num_tokens": 219230041.0, "step": 5745 }, { "epoch": 0.7309502607810711, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.756412506103516, "learning_rate": 1e-06, "loss": 0.4975, "mean_token_accuracy": 0.839177668094635, "num_tokens": 219262748.0, "step": 5746 }, { "epoch": 0.7310774710596616, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.810284614562988, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8723923563957214, "num_tokens": 219296556.0, "step": 5747 }, { "epoch": 0.7312046813382521, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.786645889282227, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.847642183303833, "num_tokens": 219330801.0, "step": 5748 }, { "epoch": 0.7313318916168426, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.724278450012207, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8617058992385864, "num_tokens": 219367470.0, "step": 5749 }, { "epoch": 0.7314591018954332, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.747774124145508, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8472216129302979, "num_tokens": 219405291.0, "step": 5750 }, { "epoch": 0.7315863121740237, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.684296607971191, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8623270392417908, "num_tokens": 219443365.0, "step": 5751 }, { "epoch": 0.7317135224526141, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.749707221984863, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8625712990760803, "num_tokens": 219477524.0, "step": 5752 }, { "epoch": 0.7318407327312046, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.691536903381348, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8525964021682739, "num_tokens": 219514576.0, "step": 5753 }, { "epoch": 0.7319679430097952, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.787132263183594, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8612418174743652, "num_tokens": 219558253.0, "step": 5754 }, { "epoch": 0.7320951532883857, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.666271209716797, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8564729690551758, "num_tokens": 219601156.0, "step": 5755 }, { "epoch": 0.7322223635669762, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.737448692321777, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8498594164848328, "num_tokens": 219638241.0, "step": 5756 }, { "epoch": 0.7323495738455668, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.739441871643066, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8616835474967957, "num_tokens": 219678168.0, "step": 5757 }, { "epoch": 0.7324767841241572, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.708663940429688, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8580408692359924, "num_tokens": 219718115.0, "step": 5758 }, { "epoch": 0.7326039944027477, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.715964317321777, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8603699207305908, "num_tokens": 219754221.0, "step": 5759 }, { "epoch": 0.7327312046813382, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.728730201721191, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8664812445640564, "num_tokens": 219786880.0, "step": 5760 }, { "epoch": 0.7328584149599288, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.721456527709961, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8642116785049438, "num_tokens": 219830585.0, "step": 5761 }, { "epoch": 0.7329856252385193, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.701740264892578, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.864536702632904, "num_tokens": 219867249.0, "step": 5762 }, { "epoch": 0.7331128355171098, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.777402877807617, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8528000116348267, "num_tokens": 219906046.0, "step": 5763 }, { "epoch": 0.7332400457957003, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.702037811279297, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8514891266822815, "num_tokens": 219947466.0, "step": 5764 }, { "epoch": 0.7333672560742908, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.703977584838867, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8629285097122192, "num_tokens": 219985698.0, "step": 5765 }, { "epoch": 0.7334944663528813, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.743494033813477, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8611627817153931, "num_tokens": 220026158.0, "step": 5766 }, { "epoch": 0.7336216766314718, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.69088363647461, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8585662245750427, "num_tokens": 220071849.0, "step": 5767 }, { "epoch": 0.7337488869100623, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.722528457641602, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8674318790435791, "num_tokens": 220110255.0, "step": 5768 }, { "epoch": 0.7338760971886529, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.758987426757812, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.8463765382766724, "num_tokens": 220144962.0, "step": 5769 }, { "epoch": 0.7340033074672434, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.742351531982422, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.851703405380249, "num_tokens": 220182118.0, "step": 5770 }, { "epoch": 0.7341305177458338, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.74522876739502, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8522869944572449, "num_tokens": 220218954.0, "step": 5771 }, { "epoch": 0.7342577280244243, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.765889167785645, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.859028697013855, "num_tokens": 220259116.0, "step": 5772 }, { "epoch": 0.7343849383030149, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.73621940612793, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8573704957962036, "num_tokens": 220301742.0, "step": 5773 }, { "epoch": 0.7345121485816054, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.719489097595215, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8645182251930237, "num_tokens": 220342572.0, "step": 5774 }, { "epoch": 0.7346393588601959, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.731498718261719, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8667088150978088, "num_tokens": 220376373.0, "step": 5775 }, { "epoch": 0.7347665691387865, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.695844650268555, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8765703439712524, "num_tokens": 220413290.0, "step": 5776 }, { "epoch": 0.7348937794173769, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.752937316894531, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8570296764373779, "num_tokens": 220453219.0, "step": 5777 }, { "epoch": 0.7350209896959674, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.754405975341797, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8613042831420898, "num_tokens": 220488369.0, "step": 5778 }, { "epoch": 0.7351481999745579, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.74246597290039, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8646706938743591, "num_tokens": 220527798.0, "step": 5779 }, { "epoch": 0.7352754102531485, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.726290702819824, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8786129951477051, "num_tokens": 220565398.0, "step": 5780 }, { "epoch": 0.735402620531739, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.77270221710205, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8696259260177612, "num_tokens": 220603903.0, "step": 5781 }, { "epoch": 0.7355298308103295, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.687925338745117, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8425431251525879, "num_tokens": 220644168.0, "step": 5782 }, { "epoch": 0.7356570410889199, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.763840675354004, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8734289407730103, "num_tokens": 220679709.0, "step": 5783 }, { "epoch": 0.7357842513675105, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.679786682128906, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.855457067489624, "num_tokens": 220721184.0, "step": 5784 }, { "epoch": 0.735911461646101, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.728716850280762, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8654960989952087, "num_tokens": 220755027.0, "step": 5785 }, { "epoch": 0.7360386719246915, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.75208568572998, "learning_rate": 1e-06, "loss": 0.5039, "mean_token_accuracy": 0.8367658257484436, "num_tokens": 220793597.0, "step": 5786 }, { "epoch": 0.736165882203282, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.741822242736816, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8597500920295715, "num_tokens": 220833549.0, "step": 5787 }, { "epoch": 0.7362930924818726, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.744609832763672, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8665557503700256, "num_tokens": 220865551.0, "step": 5788 }, { "epoch": 0.736420302760463, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.753665924072266, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8721758127212524, "num_tokens": 220911626.0, "step": 5789 }, { "epoch": 0.7365475130390535, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.749923706054688, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8618865013122559, "num_tokens": 220949839.0, "step": 5790 }, { "epoch": 0.736674723317644, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.79832649230957, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8447095155715942, "num_tokens": 220984069.0, "step": 5791 }, { "epoch": 0.7368019335962346, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.722946166992188, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8809412717819214, "num_tokens": 221020999.0, "step": 5792 }, { "epoch": 0.7369291438748251, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.781107902526855, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8508564829826355, "num_tokens": 221051082.0, "step": 5793 }, { "epoch": 0.7370563541534156, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.743520736694336, "learning_rate": 1e-06, "loss": 0.4906, "mean_token_accuracy": 0.843379557132721, "num_tokens": 221095444.0, "step": 5794 }, { "epoch": 0.737183564432006, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.816429138183594, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8629195690155029, "num_tokens": 221136433.0, "step": 5795 }, { "epoch": 0.7373107747105966, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.751810073852539, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8534296154975891, "num_tokens": 221171278.0, "step": 5796 }, { "epoch": 0.7374379849891871, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.726365089416504, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8595659136772156, "num_tokens": 221209380.0, "step": 5797 }, { "epoch": 0.7375651952677776, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.698359489440918, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8590781688690186, "num_tokens": 221253076.0, "step": 5798 }, { "epoch": 0.7376924055463682, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.763269424438477, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8564833402633667, "num_tokens": 221290348.0, "step": 5799 }, { "epoch": 0.7378196158249587, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.729065895080566, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8669115304946899, "num_tokens": 221329635.0, "step": 5800 }, { "epoch": 0.7379468261035491, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.72639274597168, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8672020435333252, "num_tokens": 221368521.0, "step": 5801 }, { "epoch": 0.7380740363821396, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.77734088897705, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8725279569625854, "num_tokens": 221407090.0, "step": 5802 }, { "epoch": 0.7382012466607302, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.697213172912598, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8695691823959351, "num_tokens": 221450340.0, "step": 5803 }, { "epoch": 0.7383284569393207, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.744441986083984, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8607029914855957, "num_tokens": 221491996.0, "step": 5804 }, { "epoch": 0.7384556672179112, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.722614288330078, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8699057102203369, "num_tokens": 221531466.0, "step": 5805 }, { "epoch": 0.7385828774965018, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.808640480041504, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8412215709686279, "num_tokens": 221567003.0, "step": 5806 }, { "epoch": 0.7387100877750922, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.78446102142334, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8734173774719238, "num_tokens": 221607396.0, "step": 5807 }, { "epoch": 0.7388372980536827, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.789356231689453, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8538577556610107, "num_tokens": 221644776.0, "step": 5808 }, { "epoch": 0.7389645083322732, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.821380615234375, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8725346326828003, "num_tokens": 221685643.0, "step": 5809 }, { "epoch": 0.7390917186108638, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.838515281677246, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8536021113395691, "num_tokens": 221724673.0, "step": 5810 }, { "epoch": 0.7392189288894543, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.70370101928711, "learning_rate": 1e-06, "loss": 0.5162, "mean_token_accuracy": 0.8371777534484863, "num_tokens": 221766287.0, "step": 5811 }, { "epoch": 0.7393461391680448, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.821328163146973, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.860004186630249, "num_tokens": 221805357.0, "step": 5812 }, { "epoch": 0.7394733494466353, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.756516456604004, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8583946824073792, "num_tokens": 221838513.0, "step": 5813 }, { "epoch": 0.7396005597252258, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.778271675109863, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8503357172012329, "num_tokens": 221882808.0, "step": 5814 }, { "epoch": 0.7397277700038163, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.75578498840332, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8736412525177002, "num_tokens": 221918505.0, "step": 5815 }, { "epoch": 0.7398549802824068, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.757624626159668, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8618521094322205, "num_tokens": 221955983.0, "step": 5816 }, { "epoch": 0.7399821905609973, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.7716064453125, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8599851131439209, "num_tokens": 221997363.0, "step": 5817 }, { "epoch": 0.7401094008395879, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.754794120788574, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8457931876182556, "num_tokens": 222039302.0, "step": 5818 }, { "epoch": 0.7402366111181784, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.785662651062012, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8602168560028076, "num_tokens": 222075779.0, "step": 5819 }, { "epoch": 0.7403638213967688, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.742162704467773, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8796850442886353, "num_tokens": 222115120.0, "step": 5820 }, { "epoch": 0.7404910316753593, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.77266788482666, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8520610332489014, "num_tokens": 222156570.0, "step": 5821 }, { "epoch": 0.7406182419539499, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.749704360961914, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8602274060249329, "num_tokens": 222189487.0, "step": 5822 }, { "epoch": 0.7407454522325404, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.781847953796387, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8592398166656494, "num_tokens": 222223163.0, "step": 5823 }, { "epoch": 0.7408726625111309, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.738906860351562, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8735237121582031, "num_tokens": 222259042.0, "step": 5824 }, { "epoch": 0.7409998727897215, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.78426742553711, "learning_rate": 1e-06, "loss": 0.5064, "mean_token_accuracy": 0.8414962291717529, "num_tokens": 222300497.0, "step": 5825 }, { "epoch": 0.7411270830683119, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.75104808807373, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.856512188911438, "num_tokens": 222333630.0, "step": 5826 }, { "epoch": 0.7412542933469024, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.733245849609375, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8560481667518616, "num_tokens": 222371715.0, "step": 5827 }, { "epoch": 0.7413815036254929, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.76227855682373, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8618950843811035, "num_tokens": 222412044.0, "step": 5828 }, { "epoch": 0.7415087139040835, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.74920654296875, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.863211989402771, "num_tokens": 222452668.0, "step": 5829 }, { "epoch": 0.741635924182674, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.841474533081055, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8620473146438599, "num_tokens": 222485631.0, "step": 5830 }, { "epoch": 0.7417631344612645, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.725052833557129, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8754920959472656, "num_tokens": 222522967.0, "step": 5831 }, { "epoch": 0.7418903447398549, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.736649513244629, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8502388000488281, "num_tokens": 222563258.0, "step": 5832 }, { "epoch": 0.7420175550184455, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.764915466308594, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8542975783348083, "num_tokens": 222599248.0, "step": 5833 }, { "epoch": 0.742144765297036, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.690664291381836, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8673075437545776, "num_tokens": 222642575.0, "step": 5834 }, { "epoch": 0.7422719755756265, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.702478408813477, "learning_rate": 1e-06, "loss": 0.5014, "mean_token_accuracy": 0.8379350900650024, "num_tokens": 222687330.0, "step": 5835 }, { "epoch": 0.742399185854217, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.755572319030762, "learning_rate": 1e-06, "loss": 0.5165, "mean_token_accuracy": 0.8353128433227539, "num_tokens": 222720020.0, "step": 5836 }, { "epoch": 0.7425263961328076, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.733634948730469, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8741660714149475, "num_tokens": 222758634.0, "step": 5837 }, { "epoch": 0.742653606411398, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.81797981262207, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8673585653305054, "num_tokens": 222794823.0, "step": 5838 }, { "epoch": 0.7427808166899885, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.735763549804688, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8601233959197998, "num_tokens": 222833836.0, "step": 5839 }, { "epoch": 0.742908026968579, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.783780097961426, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8612620830535889, "num_tokens": 222875712.0, "step": 5840 }, { "epoch": 0.7430352372471696, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.748552322387695, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8505532741546631, "num_tokens": 222915494.0, "step": 5841 }, { "epoch": 0.7431624475257601, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.73094367980957, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8773272037506104, "num_tokens": 222951675.0, "step": 5842 }, { "epoch": 0.7432896578043506, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.745682716369629, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8551132082939148, "num_tokens": 222993245.0, "step": 5843 }, { "epoch": 0.743416868082941, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.681349754333496, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8666708469390869, "num_tokens": 223029876.0, "step": 5844 }, { "epoch": 0.7435440783615316, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.757041931152344, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8658624887466431, "num_tokens": 223072550.0, "step": 5845 }, { "epoch": 0.7436712886401221, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.664300918579102, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8591195344924927, "num_tokens": 223112892.0, "step": 5846 }, { "epoch": 0.7437984989187126, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.692638397216797, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8660048246383667, "num_tokens": 223157093.0, "step": 5847 }, { "epoch": 0.7439257091973032, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.75391960144043, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8520137667655945, "num_tokens": 223190756.0, "step": 5848 }, { "epoch": 0.7440529194758937, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.78415298461914, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.8523699641227722, "num_tokens": 223226013.0, "step": 5849 }, { "epoch": 0.7441801297544841, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.719107627868652, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8738617897033691, "num_tokens": 223262858.0, "step": 5850 }, { "epoch": 0.7443073400330746, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.77039623260498, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8647279143333435, "num_tokens": 223310528.0, "step": 5851 }, { "epoch": 0.7444345503116652, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.711509704589844, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8627977967262268, "num_tokens": 223345021.0, "step": 5852 }, { "epoch": 0.7445617605902557, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.807982444763184, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8707501888275146, "num_tokens": 223382055.0, "step": 5853 }, { "epoch": 0.7446889708688462, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.765721321105957, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.850517988204956, "num_tokens": 223418441.0, "step": 5854 }, { "epoch": 0.7448161811474368, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.774816513061523, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8589460253715515, "num_tokens": 223458150.0, "step": 5855 }, { "epoch": 0.7449433914260272, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.762727737426758, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8565258383750916, "num_tokens": 223496916.0, "step": 5856 }, { "epoch": 0.7450706017046177, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.772072792053223, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8666918277740479, "num_tokens": 223534808.0, "step": 5857 }, { "epoch": 0.7451978119832082, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.74766731262207, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8703713417053223, "num_tokens": 223571255.0, "step": 5858 }, { "epoch": 0.7453250222617988, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.717059135437012, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8635742664337158, "num_tokens": 223616608.0, "step": 5859 }, { "epoch": 0.7454522325403893, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.792915344238281, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8606204986572266, "num_tokens": 223656188.0, "step": 5860 }, { "epoch": 0.7455794428189798, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.791879653930664, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8555554747581482, "num_tokens": 223693787.0, "step": 5861 }, { "epoch": 0.7457066530975703, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.758580207824707, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8726569414138794, "num_tokens": 223729350.0, "step": 5862 }, { "epoch": 0.7458338633761608, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.74255084991455, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8490229845046997, "num_tokens": 223770846.0, "step": 5863 }, { "epoch": 0.7459610736547513, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.702751159667969, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8606140613555908, "num_tokens": 223814194.0, "step": 5864 }, { "epoch": 0.7460882839333418, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.806331634521484, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8540628552436829, "num_tokens": 223849141.0, "step": 5865 }, { "epoch": 0.7462154942119323, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.764618873596191, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8815199136734009, "num_tokens": 223888654.0, "step": 5866 }, { "epoch": 0.7463427044905229, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.733792304992676, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8646633625030518, "num_tokens": 223924509.0, "step": 5867 }, { "epoch": 0.7464699147691134, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.820865631103516, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8465272188186646, "num_tokens": 223966191.0, "step": 5868 }, { "epoch": 0.7465971250477038, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.856928825378418, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8629879951477051, "num_tokens": 224001415.0, "step": 5869 }, { "epoch": 0.7467243353262943, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.799991607666016, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8675656318664551, "num_tokens": 224042820.0, "step": 5870 }, { "epoch": 0.7468515456048849, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.783493995666504, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.865885853767395, "num_tokens": 224078411.0, "step": 5871 }, { "epoch": 0.7469787558834754, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.741600036621094, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8669414520263672, "num_tokens": 224117136.0, "step": 5872 }, { "epoch": 0.7471059661620659, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.851052284240723, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8548479080200195, "num_tokens": 224155879.0, "step": 5873 }, { "epoch": 0.7472331764406565, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.789731979370117, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8723195791244507, "num_tokens": 224195186.0, "step": 5874 }, { "epoch": 0.7473603867192469, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.703359603881836, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.858667254447937, "num_tokens": 224235722.0, "step": 5875 }, { "epoch": 0.7474875969978374, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.815873146057129, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.86869215965271, "num_tokens": 224268461.0, "step": 5876 }, { "epoch": 0.7476148072764279, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.768694877624512, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.874106228351593, "num_tokens": 224302921.0, "step": 5877 }, { "epoch": 0.7477420175550185, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.803298950195312, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.862012505531311, "num_tokens": 224344081.0, "step": 5878 }, { "epoch": 0.747869227833609, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.72903823852539, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8500140309333801, "num_tokens": 224384464.0, "step": 5879 }, { "epoch": 0.7479964381121995, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.785969734191895, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8696489930152893, "num_tokens": 224425542.0, "step": 5880 }, { "epoch": 0.7481236483907899, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.740070343017578, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8777472972869873, "num_tokens": 224465342.0, "step": 5881 }, { "epoch": 0.7482508586693805, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.834945678710938, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8665745854377747, "num_tokens": 224504787.0, "step": 5882 }, { "epoch": 0.748378068947971, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.722479820251465, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8618361949920654, "num_tokens": 224548534.0, "step": 5883 }, { "epoch": 0.7485052792265615, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.816264152526855, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.852832555770874, "num_tokens": 224590251.0, "step": 5884 }, { "epoch": 0.748632489505152, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.747167587280273, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8717295527458191, "num_tokens": 224633745.0, "step": 5885 }, { "epoch": 0.7487596997837426, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.785258293151855, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8464272022247314, "num_tokens": 224671319.0, "step": 5886 }, { "epoch": 0.748886910062333, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.824211120605469, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8615070581436157, "num_tokens": 224712670.0, "step": 5887 }, { "epoch": 0.7490141203409235, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.742233276367188, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8594468235969543, "num_tokens": 224747840.0, "step": 5888 }, { "epoch": 0.749141330619514, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.8065185546875, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8709858655929565, "num_tokens": 224783274.0, "step": 5889 }, { "epoch": 0.7492685408981046, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.788467407226562, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8565695285797119, "num_tokens": 224820456.0, "step": 5890 }, { "epoch": 0.7493957511766951, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.811582565307617, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8656922578811646, "num_tokens": 224858569.0, "step": 5891 }, { "epoch": 0.7495229614552856, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.81832218170166, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8594701886177063, "num_tokens": 224901033.0, "step": 5892 }, { "epoch": 0.749650171733876, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.762651443481445, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8765006065368652, "num_tokens": 224934345.0, "step": 5893 }, { "epoch": 0.7497773820124666, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.836817741394043, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8546308279037476, "num_tokens": 224972479.0, "step": 5894 }, { "epoch": 0.7499045922910571, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.77949333190918, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8787516951560974, "num_tokens": 225014213.0, "step": 5895 }, { "epoch": 0.7500318025696476, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.750051498413086, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8678830862045288, "num_tokens": 225049929.0, "step": 5896 }, { "epoch": 0.7501590128482382, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.857000350952148, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.845142662525177, "num_tokens": 225086409.0, "step": 5897 }, { "epoch": 0.7502862231268287, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.827953338623047, "learning_rate": 1e-06, "loss": 0.4903, "mean_token_accuracy": 0.8436497449874878, "num_tokens": 225124443.0, "step": 5898 }, { "epoch": 0.7504134334054191, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.839683532714844, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8508838415145874, "num_tokens": 225166665.0, "step": 5899 }, { "epoch": 0.7505406436840096, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.79240608215332, "learning_rate": 1e-06, "loss": 0.5116, "mean_token_accuracy": 0.8373960852622986, "num_tokens": 225203841.0, "step": 5900 }, { "epoch": 0.7506678539626002, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.857071876525879, "learning_rate": 1e-06, "loss": 0.4816, "mean_token_accuracy": 0.8446094989776611, "num_tokens": 225234844.0, "step": 5901 }, { "epoch": 0.7507950642411907, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.837507247924805, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.842762291431427, "num_tokens": 225269828.0, "step": 5902 }, { "epoch": 0.7509222745197812, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.788228034973145, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.86576908826828, "num_tokens": 225312361.0, "step": 5903 }, { "epoch": 0.7510494847983717, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.831147193908691, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.8509451150894165, "num_tokens": 225353708.0, "step": 5904 }, { "epoch": 0.7511766950769622, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.762686729431152, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8800584077835083, "num_tokens": 225399730.0, "step": 5905 }, { "epoch": 0.7513039053555527, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.882378578186035, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8550246953964233, "num_tokens": 225441836.0, "step": 5906 }, { "epoch": 0.7514311156341432, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.783307075500488, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8682661056518555, "num_tokens": 225471004.0, "step": 5907 }, { "epoch": 0.7515583259127337, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.897428512573242, "learning_rate": 1e-06, "loss": 0.4895, "mean_token_accuracy": 0.8428245782852173, "num_tokens": 225505789.0, "step": 5908 }, { "epoch": 0.7516855361913243, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.768290519714355, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8633471727371216, "num_tokens": 225541138.0, "step": 5909 }, { "epoch": 0.7518127464699148, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.824161529541016, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8644037246704102, "num_tokens": 225581695.0, "step": 5910 }, { "epoch": 0.7519399567485053, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.794088363647461, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8664454221725464, "num_tokens": 225619532.0, "step": 5911 }, { "epoch": 0.7520671670270958, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.780515670776367, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8524743914604187, "num_tokens": 225657403.0, "step": 5912 }, { "epoch": 0.7521943773056863, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.82028865814209, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8568543195724487, "num_tokens": 225693402.0, "step": 5913 }, { "epoch": 0.7523215875842768, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.861063957214355, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8596492409706116, "num_tokens": 225727208.0, "step": 5914 }, { "epoch": 0.7524487978628673, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.818528175354004, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8655878901481628, "num_tokens": 225762117.0, "step": 5915 }, { "epoch": 0.7525760081414579, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.748015403747559, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8465303182601929, "num_tokens": 225794737.0, "step": 5916 }, { "epoch": 0.7527032184200484, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.803121566772461, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8595547080039978, "num_tokens": 225829074.0, "step": 5917 }, { "epoch": 0.7528304286986388, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.772329330444336, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8742634654045105, "num_tokens": 225871344.0, "step": 5918 }, { "epoch": 0.7529576389772293, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.81345272064209, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8560280203819275, "num_tokens": 225911321.0, "step": 5919 }, { "epoch": 0.7530848492558199, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.796662330627441, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8699381947517395, "num_tokens": 225944613.0, "step": 5920 }, { "epoch": 0.7532120595344104, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.820627212524414, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8657673597335815, "num_tokens": 225989501.0, "step": 5921 }, { "epoch": 0.7533392698130009, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.772945404052734, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8617143630981445, "num_tokens": 226024984.0, "step": 5922 }, { "epoch": 0.7534664800915915, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.782876968383789, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8737828731536865, "num_tokens": 226063255.0, "step": 5923 }, { "epoch": 0.7535936903701819, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.834174156188965, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8486287593841553, "num_tokens": 226105538.0, "step": 5924 }, { "epoch": 0.7537209006487724, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.847186088562012, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8645542860031128, "num_tokens": 226137449.0, "step": 5925 }, { "epoch": 0.7538481109273629, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.830536842346191, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8549291491508484, "num_tokens": 226171078.0, "step": 5926 }, { "epoch": 0.7539753212059535, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.797248840332031, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8626827597618103, "num_tokens": 226208060.0, "step": 5927 }, { "epoch": 0.754102531484544, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.82666015625, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8705960512161255, "num_tokens": 226245330.0, "step": 5928 }, { "epoch": 0.7542297417631345, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.793513298034668, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8827273845672607, "num_tokens": 226281974.0, "step": 5929 }, { "epoch": 0.7543569520417249, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.779007911682129, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.850804328918457, "num_tokens": 226323302.0, "step": 5930 }, { "epoch": 0.7544841623203155, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.815611839294434, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8485901355743408, "num_tokens": 226361346.0, "step": 5931 }, { "epoch": 0.754611372598906, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.809410095214844, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.856904923915863, "num_tokens": 226397744.0, "step": 5932 }, { "epoch": 0.7547385828774965, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.81507682800293, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8687975406646729, "num_tokens": 226436745.0, "step": 5933 }, { "epoch": 0.754865793156087, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.783868789672852, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8692160248756409, "num_tokens": 226470302.0, "step": 5934 }, { "epoch": 0.7549930034346776, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.816201210021973, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8641221523284912, "num_tokens": 226500090.0, "step": 5935 }, { "epoch": 0.755120213713268, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.874865531921387, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8618786931037903, "num_tokens": 226533359.0, "step": 5936 }, { "epoch": 0.7552474239918585, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.807906150817871, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8521504402160645, "num_tokens": 226567060.0, "step": 5937 }, { "epoch": 0.755374634270449, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.81928825378418, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8595982789993286, "num_tokens": 226604270.0, "step": 5938 }, { "epoch": 0.7555018445490396, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.782585144042969, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8591926097869873, "num_tokens": 226641310.0, "step": 5939 }, { "epoch": 0.7556290548276301, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.768421173095703, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8509781360626221, "num_tokens": 226678305.0, "step": 5940 }, { "epoch": 0.7557562651062206, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.832318305969238, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8684045076370239, "num_tokens": 226715628.0, "step": 5941 }, { "epoch": 0.755883475384811, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.758418083190918, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8725530505180359, "num_tokens": 226755346.0, "step": 5942 }, { "epoch": 0.7560106856634016, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.874137878417969, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8506664037704468, "num_tokens": 226792243.0, "step": 5943 }, { "epoch": 0.7561378959419921, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.763323783874512, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8752095699310303, "num_tokens": 226829033.0, "step": 5944 }, { "epoch": 0.7562651062205826, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.819538116455078, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8758481740951538, "num_tokens": 226870645.0, "step": 5945 }, { "epoch": 0.7563923164991732, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.808267593383789, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8652739524841309, "num_tokens": 226902395.0, "step": 5946 }, { "epoch": 0.7565195267777637, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.848508834838867, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8679736852645874, "num_tokens": 226942620.0, "step": 5947 }, { "epoch": 0.7566467370563541, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.773531913757324, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8684267997741699, "num_tokens": 226976472.0, "step": 5948 }, { "epoch": 0.7567739473349446, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.819478034973145, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8513969779014587, "num_tokens": 227013626.0, "step": 5949 }, { "epoch": 0.7569011576135352, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.783339500427246, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8486562967300415, "num_tokens": 227049336.0, "step": 5950 }, { "epoch": 0.7570283678921257, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.806628227233887, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8839800953865051, "num_tokens": 227087371.0, "step": 5951 }, { "epoch": 0.7571555781707162, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.738736152648926, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8639345169067383, "num_tokens": 227129188.0, "step": 5952 }, { "epoch": 0.7572827884493067, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.80806827545166, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8553504943847656, "num_tokens": 227161807.0, "step": 5953 }, { "epoch": 0.7574099987278972, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.842824935913086, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8531792759895325, "num_tokens": 227198431.0, "step": 5954 }, { "epoch": 0.7575372090064877, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.80507755279541, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8816776275634766, "num_tokens": 227236046.0, "step": 5955 }, { "epoch": 0.7576644192850782, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.781182289123535, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8767299652099609, "num_tokens": 227272925.0, "step": 5956 }, { "epoch": 0.7577916295636687, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.811421394348145, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8658077120780945, "num_tokens": 227312311.0, "step": 5957 }, { "epoch": 0.7579188398422593, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.82624340057373, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8815221786499023, "num_tokens": 227349632.0, "step": 5958 }, { "epoch": 0.7580460501208498, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.81202507019043, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8607122898101807, "num_tokens": 227392982.0, "step": 5959 }, { "epoch": 0.7581732603994403, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.849442481994629, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.858784556388855, "num_tokens": 227431878.0, "step": 5960 }, { "epoch": 0.7583004706780307, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.830236434936523, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8522375822067261, "num_tokens": 227470401.0, "step": 5961 }, { "epoch": 0.7584276809566213, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.886481285095215, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8669828176498413, "num_tokens": 227507189.0, "step": 5962 }, { "epoch": 0.7585548912352118, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.789861679077148, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8665974140167236, "num_tokens": 227548070.0, "step": 5963 }, { "epoch": 0.7586821015138023, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.831801414489746, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8667997717857361, "num_tokens": 227584360.0, "step": 5964 }, { "epoch": 0.7588093117923929, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.817435264587402, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8685890436172485, "num_tokens": 227618961.0, "step": 5965 }, { "epoch": 0.7589365220709834, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.771533012390137, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.86301589012146, "num_tokens": 227658304.0, "step": 5966 }, { "epoch": 0.7590637323495738, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.812655448913574, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8604936599731445, "num_tokens": 227701526.0, "step": 5967 }, { "epoch": 0.7591909426281643, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.74638557434082, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8672953248023987, "num_tokens": 227744930.0, "step": 5968 }, { "epoch": 0.7593181529067549, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.863739013671875, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8732223510742188, "num_tokens": 227780607.0, "step": 5969 }, { "epoch": 0.7594453631853454, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.848026275634766, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8588782548904419, "num_tokens": 227825742.0, "step": 5970 }, { "epoch": 0.7595725734639359, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.880980491638184, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8436727523803711, "num_tokens": 227865278.0, "step": 5971 }, { "epoch": 0.7596997837425264, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.864351272583008, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8588760495185852, "num_tokens": 227906912.0, "step": 5972 }, { "epoch": 0.7598269940211169, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.89551067352295, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8602584004402161, "num_tokens": 227947067.0, "step": 5973 }, { "epoch": 0.7599542042997074, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.852021217346191, "learning_rate": 1e-06, "loss": 0.4864, "mean_token_accuracy": 0.8440934419631958, "num_tokens": 227988563.0, "step": 5974 }, { "epoch": 0.7600814145782979, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.82408618927002, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8552651405334473, "num_tokens": 228028556.0, "step": 5975 }, { "epoch": 0.7602086248568884, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.831303596496582, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8767502307891846, "num_tokens": 228066265.0, "step": 5976 }, { "epoch": 0.760335835135479, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.88369369506836, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8588189482688904, "num_tokens": 228103732.0, "step": 5977 }, { "epoch": 0.7604630454140695, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.818197250366211, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.86012202501297, "num_tokens": 228149609.0, "step": 5978 }, { "epoch": 0.7605902556926599, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.867252349853516, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8641636371612549, "num_tokens": 228188462.0, "step": 5979 }, { "epoch": 0.7607174659712505, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.827406883239746, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8757942914962769, "num_tokens": 228229252.0, "step": 5980 }, { "epoch": 0.760844676249841, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.881552696228027, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8603774309158325, "num_tokens": 228267397.0, "step": 5981 }, { "epoch": 0.7609718865284315, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.778346061706543, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8705337643623352, "num_tokens": 228307561.0, "step": 5982 }, { "epoch": 0.761099096807022, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.823546409606934, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8639246821403503, "num_tokens": 228347667.0, "step": 5983 }, { "epoch": 0.7612263070856126, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.861233711242676, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.859691321849823, "num_tokens": 228386820.0, "step": 5984 }, { "epoch": 0.761353517364203, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.824621200561523, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8715640902519226, "num_tokens": 228427304.0, "step": 5985 }, { "epoch": 0.7614807276427935, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.886368751525879, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8526928424835205, "num_tokens": 228470779.0, "step": 5986 }, { "epoch": 0.761607937921384, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.90720272064209, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8789968490600586, "num_tokens": 228500454.0, "step": 5987 }, { "epoch": 0.7617351481999746, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.847033500671387, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8694511651992798, "num_tokens": 228537374.0, "step": 5988 }, { "epoch": 0.7618623584785651, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.838485717773438, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8629225492477417, "num_tokens": 228577215.0, "step": 5989 }, { "epoch": 0.7619895687571556, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.875044822692871, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8670586347579956, "num_tokens": 228611693.0, "step": 5990 }, { "epoch": 0.762116779035746, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.84453010559082, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8613438010215759, "num_tokens": 228644231.0, "step": 5991 }, { "epoch": 0.7622439893143366, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.833108901977539, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8732838034629822, "num_tokens": 228684776.0, "step": 5992 }, { "epoch": 0.7623711995929271, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.773147583007812, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8686537742614746, "num_tokens": 228724214.0, "step": 5993 }, { "epoch": 0.7624984098715176, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.809538841247559, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8607649207115173, "num_tokens": 228765098.0, "step": 5994 }, { "epoch": 0.7626256201501082, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.841400146484375, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8640211820602417, "num_tokens": 228802423.0, "step": 5995 }, { "epoch": 0.7627528304286987, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.897032737731934, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8524613380432129, "num_tokens": 228840218.0, "step": 5996 }, { "epoch": 0.7628800407072891, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.798529624938965, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8673335313796997, "num_tokens": 228882810.0, "step": 5997 }, { "epoch": 0.7630072509858796, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.833413124084473, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8695493936538696, "num_tokens": 228917664.0, "step": 5998 }, { "epoch": 0.7631344612644702, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.85045051574707, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8696237206459045, "num_tokens": 228951017.0, "step": 5999 }, { "epoch": 0.7632616715430607, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.818269729614258, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8643230199813843, "num_tokens": 228990949.0, "step": 6000 }, { "epoch": 0.7633888818216512, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.828598022460938, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8678169250488281, "num_tokens": 229026604.0, "step": 6001 }, { "epoch": 0.7635160921002417, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.848719596862793, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8563864231109619, "num_tokens": 229061810.0, "step": 6002 }, { "epoch": 0.7636433023788322, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.84683609008789, "learning_rate": 1e-06, "loss": 0.4948, "mean_token_accuracy": 0.8400386571884155, "num_tokens": 229107765.0, "step": 6003 }, { "epoch": 0.7637705126574227, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.874850273132324, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.850103497505188, "num_tokens": 229147822.0, "step": 6004 }, { "epoch": 0.7638977229360132, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.845146179199219, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8725378513336182, "num_tokens": 229184421.0, "step": 6005 }, { "epoch": 0.7640249332146037, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.838501930236816, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8631569147109985, "num_tokens": 229221737.0, "step": 6006 }, { "epoch": 0.7641521434931943, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.822738647460938, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8571269512176514, "num_tokens": 229254727.0, "step": 6007 }, { "epoch": 0.7642793537717848, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.8339262008667, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8734917640686035, "num_tokens": 229292648.0, "step": 6008 }, { "epoch": 0.7644065640503753, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.76049518585205, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8679780960083008, "num_tokens": 229335708.0, "step": 6009 }, { "epoch": 0.7645337743289657, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.826383590698242, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.871318519115448, "num_tokens": 229373791.0, "step": 6010 }, { "epoch": 0.7646609846075563, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.852179527282715, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.873037576675415, "num_tokens": 229409891.0, "step": 6011 }, { "epoch": 0.7647881948861468, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.824853897094727, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8647546768188477, "num_tokens": 229449859.0, "step": 6012 }, { "epoch": 0.7649154051647373, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.799671173095703, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8697878122329712, "num_tokens": 229484224.0, "step": 6013 }, { "epoch": 0.7650426154433279, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.822094917297363, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8591823577880859, "num_tokens": 229521104.0, "step": 6014 }, { "epoch": 0.7651698257219184, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.840583801269531, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8479141592979431, "num_tokens": 229558918.0, "step": 6015 }, { "epoch": 0.7652970360005088, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.848292350769043, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8497313261032104, "num_tokens": 229596931.0, "step": 6016 }, { "epoch": 0.7654242462790993, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.815142631530762, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8485155701637268, "num_tokens": 229633903.0, "step": 6017 }, { "epoch": 0.7655514565576899, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.838839530944824, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8600605726242065, "num_tokens": 229674224.0, "step": 6018 }, { "epoch": 0.7656786668362804, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.794437408447266, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.861786961555481, "num_tokens": 229710547.0, "step": 6019 }, { "epoch": 0.7658058771148709, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.831881523132324, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8778102397918701, "num_tokens": 229752020.0, "step": 6020 }, { "epoch": 0.7659330873934614, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.815020561218262, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8596550822257996, "num_tokens": 229785710.0, "step": 6021 }, { "epoch": 0.7660602976720519, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.826807022094727, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8697375655174255, "num_tokens": 229824118.0, "step": 6022 }, { "epoch": 0.7661875079506424, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.817665100097656, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8745493292808533, "num_tokens": 229854231.0, "step": 6023 }, { "epoch": 0.7663147182292329, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.85447883605957, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8529365062713623, "num_tokens": 229888102.0, "step": 6024 }, { "epoch": 0.7664419285078234, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.773134231567383, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8660668134689331, "num_tokens": 229927634.0, "step": 6025 }, { "epoch": 0.766569138786414, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.83360767364502, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8710517287254333, "num_tokens": 229966526.0, "step": 6026 }, { "epoch": 0.7666963490650045, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.814940452575684, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.872444748878479, "num_tokens": 230002761.0, "step": 6027 }, { "epoch": 0.7668235593435949, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.884261131286621, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8688601851463318, "num_tokens": 230041303.0, "step": 6028 }, { "epoch": 0.7669507696221854, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.836915016174316, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8673162460327148, "num_tokens": 230081117.0, "step": 6029 }, { "epoch": 0.767077979900776, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.875707626342773, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8578513860702515, "num_tokens": 230119411.0, "step": 6030 }, { "epoch": 0.7672051901793665, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.904051780700684, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8705212473869324, "num_tokens": 230161111.0, "step": 6031 }, { "epoch": 0.767332400457957, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.8275785446167, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8587993383407593, "num_tokens": 230196999.0, "step": 6032 }, { "epoch": 0.7674596107365476, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.872653007507324, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.865297794342041, "num_tokens": 230231495.0, "step": 6033 }, { "epoch": 0.767586821015138, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.79840087890625, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8565355539321899, "num_tokens": 230267732.0, "step": 6034 }, { "epoch": 0.7677140312937285, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.871914863586426, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.862668514251709, "num_tokens": 230297559.0, "step": 6035 }, { "epoch": 0.767841241572319, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.839096069335938, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8510259985923767, "num_tokens": 230336396.0, "step": 6036 }, { "epoch": 0.7679684518509096, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.828763008117676, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8625268340110779, "num_tokens": 230368152.0, "step": 6037 }, { "epoch": 0.7680956621295001, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.810559272766113, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8684881925582886, "num_tokens": 230408081.0, "step": 6038 }, { "epoch": 0.7682228724080906, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.849620819091797, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8560743927955627, "num_tokens": 230441691.0, "step": 6039 }, { "epoch": 0.768350082686681, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.83607006072998, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8613170385360718, "num_tokens": 230480089.0, "step": 6040 }, { "epoch": 0.7684772929652716, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.82755184173584, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8570252656936646, "num_tokens": 230518655.0, "step": 6041 }, { "epoch": 0.7686045032438621, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.87594223022461, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8659923076629639, "num_tokens": 230555717.0, "step": 6042 }, { "epoch": 0.7687317135224526, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.849967002868652, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.871681809425354, "num_tokens": 230596699.0, "step": 6043 }, { "epoch": 0.7688589238010431, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.822299003601074, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8715541362762451, "num_tokens": 230633284.0, "step": 6044 }, { "epoch": 0.7689861340796337, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.859079360961914, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8670304417610168, "num_tokens": 230673133.0, "step": 6045 }, { "epoch": 0.7691133443582241, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.89241886138916, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8639606237411499, "num_tokens": 230719003.0, "step": 6046 }, { "epoch": 0.7692405546368146, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.819175720214844, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8724881410598755, "num_tokens": 230754805.0, "step": 6047 }, { "epoch": 0.7693677649154052, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.926055908203125, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8505450487136841, "num_tokens": 230793249.0, "step": 6048 }, { "epoch": 0.7694949751939957, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.858102798461914, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8807415962219238, "num_tokens": 230826025.0, "step": 6049 }, { "epoch": 0.7696221854725862, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.82265853881836, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8752980828285217, "num_tokens": 230866615.0, "step": 6050 }, { "epoch": 0.7697493957511767, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.908384323120117, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8480658531188965, "num_tokens": 230901587.0, "step": 6051 }, { "epoch": 0.7698766060297672, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.818115234375, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8688620924949646, "num_tokens": 230944267.0, "step": 6052 }, { "epoch": 0.7700038163083577, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.870530128479004, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8538505434989929, "num_tokens": 230980022.0, "step": 6053 }, { "epoch": 0.7701310265869482, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.847841262817383, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8689155578613281, "num_tokens": 231018110.0, "step": 6054 }, { "epoch": 0.7702582368655387, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.813713073730469, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8678369522094727, "num_tokens": 231056531.0, "step": 6055 }, { "epoch": 0.7703854471441293, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.841798782348633, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8685112595558167, "num_tokens": 231096270.0, "step": 6056 }, { "epoch": 0.7705126574227198, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.789107322692871, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8564901351928711, "num_tokens": 231136300.0, "step": 6057 }, { "epoch": 0.7706398677013102, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.850240707397461, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8503739833831787, "num_tokens": 231169005.0, "step": 6058 }, { "epoch": 0.7707670779799007, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.83181381225586, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8699645400047302, "num_tokens": 231205274.0, "step": 6059 }, { "epoch": 0.7708942882584913, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.84027099609375, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8686757683753967, "num_tokens": 231239022.0, "step": 6060 }, { "epoch": 0.7710214985370818, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.787213325500488, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8454632759094238, "num_tokens": 231282952.0, "step": 6061 }, { "epoch": 0.7711487088156723, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.84597110748291, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8610646724700928, "num_tokens": 231322980.0, "step": 6062 }, { "epoch": 0.7712759190942629, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.852375984191895, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8553035259246826, "num_tokens": 231358382.0, "step": 6063 }, { "epoch": 0.7714031293728534, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.788694381713867, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8715802431106567, "num_tokens": 231401915.0, "step": 6064 }, { "epoch": 0.7715303396514438, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.838458061218262, "learning_rate": 1e-06, "loss": 0.4831, "mean_token_accuracy": 0.840533435344696, "num_tokens": 231433241.0, "step": 6065 }, { "epoch": 0.7716575499300343, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.776748657226562, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8577629923820496, "num_tokens": 231478318.0, "step": 6066 }, { "epoch": 0.7717847602086249, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.879814147949219, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.855508029460907, "num_tokens": 231518584.0, "step": 6067 }, { "epoch": 0.7719119704872154, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.791847229003906, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8625088334083557, "num_tokens": 231556689.0, "step": 6068 }, { "epoch": 0.7720391807658059, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.852141380310059, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8621869683265686, "num_tokens": 231593936.0, "step": 6069 }, { "epoch": 0.7721663910443964, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.884342193603516, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8824886679649353, "num_tokens": 231634785.0, "step": 6070 }, { "epoch": 0.7722936013229869, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.816523551940918, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8590342998504639, "num_tokens": 231675798.0, "step": 6071 }, { "epoch": 0.7724208116015774, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.883955955505371, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8610267639160156, "num_tokens": 231711611.0, "step": 6072 }, { "epoch": 0.7725480218801679, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.795085906982422, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8536919951438904, "num_tokens": 231746997.0, "step": 6073 }, { "epoch": 0.7726752321587584, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.892691612243652, "learning_rate": 1e-06, "loss": 0.4847, "mean_token_accuracy": 0.8426701426506042, "num_tokens": 231787696.0, "step": 6074 }, { "epoch": 0.772802442437349, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.81282901763916, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8549165725708008, "num_tokens": 231826127.0, "step": 6075 }, { "epoch": 0.7729296527159395, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.831796646118164, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8577685356140137, "num_tokens": 231857538.0, "step": 6076 }, { "epoch": 0.7730568629945299, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.83186149597168, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8608824610710144, "num_tokens": 231893913.0, "step": 6077 }, { "epoch": 0.7731840732731204, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.884517669677734, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8657872676849365, "num_tokens": 231934274.0, "step": 6078 }, { "epoch": 0.773311283551711, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.802536010742188, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8566691875457764, "num_tokens": 231983197.0, "step": 6079 }, { "epoch": 0.7734384938303015, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.821059226989746, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8585909605026245, "num_tokens": 232021265.0, "step": 6080 }, { "epoch": 0.773565704108892, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.855996131896973, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8609038591384888, "num_tokens": 232060992.0, "step": 6081 }, { "epoch": 0.7736929143874826, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.855239868164062, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.860167920589447, "num_tokens": 232096758.0, "step": 6082 }, { "epoch": 0.773820124666073, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.828758239746094, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8585935831069946, "num_tokens": 232133773.0, "step": 6083 }, { "epoch": 0.7739473349446635, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.801566123962402, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8609617948532104, "num_tokens": 232173182.0, "step": 6084 }, { "epoch": 0.774074545223254, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.797385215759277, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8704965114593506, "num_tokens": 232206478.0, "step": 6085 }, { "epoch": 0.7742017555018446, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.822031021118164, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.856386661529541, "num_tokens": 232242231.0, "step": 6086 }, { "epoch": 0.7743289657804351, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.880056381225586, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8658707141876221, "num_tokens": 232270636.0, "step": 6087 }, { "epoch": 0.7744561760590256, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.928693771362305, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.868008553981781, "num_tokens": 232309974.0, "step": 6088 }, { "epoch": 0.774583386337616, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.805802345275879, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.871764600276947, "num_tokens": 232344118.0, "step": 6089 }, { "epoch": 0.7747105966162066, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.851884841918945, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8519930839538574, "num_tokens": 232385183.0, "step": 6090 }, { "epoch": 0.7748378068947971, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.827469825744629, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8639041185379028, "num_tokens": 232421670.0, "step": 6091 }, { "epoch": 0.7749650171733876, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.819723129272461, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8503644466400146, "num_tokens": 232459563.0, "step": 6092 }, { "epoch": 0.7750922274519781, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.84302806854248, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8625848293304443, "num_tokens": 232496357.0, "step": 6093 }, { "epoch": 0.7752194377305687, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.878540992736816, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.862174928188324, "num_tokens": 232530436.0, "step": 6094 }, { "epoch": 0.7753466480091591, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.858516693115234, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8564088344573975, "num_tokens": 232568585.0, "step": 6095 }, { "epoch": 0.7754738582877496, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.872550964355469, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8686933517456055, "num_tokens": 232602506.0, "step": 6096 }, { "epoch": 0.7756010685663401, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.843770027160645, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8702371716499329, "num_tokens": 232642819.0, "step": 6097 }, { "epoch": 0.7757282788449307, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.833791732788086, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8494265079498291, "num_tokens": 232679505.0, "step": 6098 }, { "epoch": 0.7758554891235212, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.839591979980469, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8675315380096436, "num_tokens": 232715701.0, "step": 6099 }, { "epoch": 0.7759826994021117, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.88361930847168, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8638358116149902, "num_tokens": 232750786.0, "step": 6100 }, { "epoch": 0.7761099096807021, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.85079288482666, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8508377075195312, "num_tokens": 232791370.0, "step": 6101 }, { "epoch": 0.7762371199592927, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.785449028015137, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.86406409740448, "num_tokens": 232830895.0, "step": 6102 }, { "epoch": 0.7763643302378832, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.908076286315918, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.861191987991333, "num_tokens": 232869940.0, "step": 6103 }, { "epoch": 0.7764915405164737, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.782965660095215, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8582634925842285, "num_tokens": 232910743.0, "step": 6104 }, { "epoch": 0.7766187507950643, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.891020774841309, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8609672784805298, "num_tokens": 232948428.0, "step": 6105 }, { "epoch": 0.7767459610736548, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.888991355895996, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8682698607444763, "num_tokens": 232985142.0, "step": 6106 }, { "epoch": 0.7768731713522452, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.88514518737793, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8718231916427612, "num_tokens": 233020810.0, "step": 6107 }, { "epoch": 0.7770003816308357, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.87528133392334, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8544808626174927, "num_tokens": 233055674.0, "step": 6108 }, { "epoch": 0.7771275919094263, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.891836166381836, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8573637008666992, "num_tokens": 233084506.0, "step": 6109 }, { "epoch": 0.7772548021880168, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.910723686218262, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8785531520843506, "num_tokens": 233120172.0, "step": 6110 }, { "epoch": 0.7773820124666073, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.857351303100586, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8525592088699341, "num_tokens": 233160075.0, "step": 6111 }, { "epoch": 0.7775092227451978, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.910706520080566, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8545460104942322, "num_tokens": 233201714.0, "step": 6112 }, { "epoch": 0.7776364330237884, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.868894577026367, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8686857223510742, "num_tokens": 233237642.0, "step": 6113 }, { "epoch": 0.7777636433023788, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.852385520935059, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8673927783966064, "num_tokens": 233279712.0, "step": 6114 }, { "epoch": 0.7778908535809693, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.828394889831543, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8681949973106384, "num_tokens": 233322441.0, "step": 6115 }, { "epoch": 0.7780180638595598, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.871986389160156, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.856086015701294, "num_tokens": 233371053.0, "step": 6116 }, { "epoch": 0.7781452741381504, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.942676544189453, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.86383056640625, "num_tokens": 233405372.0, "step": 6117 }, { "epoch": 0.7782724844167409, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.879203796386719, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8689719438552856, "num_tokens": 233447209.0, "step": 6118 }, { "epoch": 0.7783996946953314, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.894817352294922, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8652521371841431, "num_tokens": 233484621.0, "step": 6119 }, { "epoch": 0.7785269049739219, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.8734769821167, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8451753854751587, "num_tokens": 233520756.0, "step": 6120 }, { "epoch": 0.7786541152525124, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.899832725524902, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8551538586616516, "num_tokens": 233559120.0, "step": 6121 }, { "epoch": 0.7787813255311029, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.825523376464844, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8738962411880493, "num_tokens": 233590513.0, "step": 6122 }, { "epoch": 0.7789085358096934, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.860617637634277, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8784643411636353, "num_tokens": 233625986.0, "step": 6123 }, { "epoch": 0.779035746088284, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.935416221618652, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8713473081588745, "num_tokens": 233655097.0, "step": 6124 }, { "epoch": 0.7791629563668745, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.820677757263184, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8525862693786621, "num_tokens": 233687815.0, "step": 6125 }, { "epoch": 0.7792901666454649, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.879538536071777, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.875998318195343, "num_tokens": 233722611.0, "step": 6126 }, { "epoch": 0.7794173769240554, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.885266304016113, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.850139856338501, "num_tokens": 233765345.0, "step": 6127 }, { "epoch": 0.779544587202646, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.841506004333496, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8641716241836548, "num_tokens": 233804358.0, "step": 6128 }, { "epoch": 0.7796717974812365, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.788773536682129, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8522123694419861, "num_tokens": 233845117.0, "step": 6129 }, { "epoch": 0.779799007759827, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.881438255310059, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8550970554351807, "num_tokens": 233884091.0, "step": 6130 }, { "epoch": 0.7799262180384176, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.767264366149902, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8599493503570557, "num_tokens": 233922590.0, "step": 6131 }, { "epoch": 0.780053428317008, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.893284797668457, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8652770519256592, "num_tokens": 233954138.0, "step": 6132 }, { "epoch": 0.7801806385955985, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.864293098449707, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8721012473106384, "num_tokens": 233988035.0, "step": 6133 }, { "epoch": 0.780307848874189, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.865920066833496, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8529462814331055, "num_tokens": 234023747.0, "step": 6134 }, { "epoch": 0.7804350591527796, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.826898574829102, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8703223466873169, "num_tokens": 234064182.0, "step": 6135 }, { "epoch": 0.7805622694313701, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.859797477722168, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.846349835395813, "num_tokens": 234106533.0, "step": 6136 }, { "epoch": 0.7806894797099606, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.890552520751953, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.874600887298584, "num_tokens": 234145492.0, "step": 6137 }, { "epoch": 0.780816689988551, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.859933853149414, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.865517258644104, "num_tokens": 234181709.0, "step": 6138 }, { "epoch": 0.7809439002671416, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.846332550048828, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8627538084983826, "num_tokens": 234217854.0, "step": 6139 }, { "epoch": 0.7810711105457321, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.861268997192383, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.86278235912323, "num_tokens": 234255769.0, "step": 6140 }, { "epoch": 0.7811983208243226, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.926581382751465, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8493065237998962, "num_tokens": 234290828.0, "step": 6141 }, { "epoch": 0.7813255311029131, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.865545272827148, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8617315888404846, "num_tokens": 234326336.0, "step": 6142 }, { "epoch": 0.7814527413815037, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.899364471435547, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8570688962936401, "num_tokens": 234366210.0, "step": 6143 }, { "epoch": 0.7815799516600941, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.803911209106445, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8543119430541992, "num_tokens": 234398975.0, "step": 6144 }, { "epoch": 0.7817071619386846, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.890666007995605, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8568556308746338, "num_tokens": 234437746.0, "step": 6145 }, { "epoch": 0.7818343722172751, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.788909912109375, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8670765161514282, "num_tokens": 234478402.0, "step": 6146 }, { "epoch": 0.7819615824958657, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.821105003356934, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8481664657592773, "num_tokens": 234509699.0, "step": 6147 }, { "epoch": 0.7820887927744562, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.828120231628418, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8567724823951721, "num_tokens": 234551515.0, "step": 6148 }, { "epoch": 0.7822160030530467, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.817850112915039, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8694463968276978, "num_tokens": 234582811.0, "step": 6149 }, { "epoch": 0.7823432133316371, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.837686538696289, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8546777963638306, "num_tokens": 234625299.0, "step": 6150 }, { "epoch": 0.7824704236102277, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.771035194396973, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8689693212509155, "num_tokens": 234665641.0, "step": 6151 }, { "epoch": 0.7825976338888182, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.874859809875488, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8621986508369446, "num_tokens": 234704668.0, "step": 6152 }, { "epoch": 0.7827248441674087, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.882099151611328, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8674322366714478, "num_tokens": 234738080.0, "step": 6153 }, { "epoch": 0.7828520544459993, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.863986015319824, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8602281808853149, "num_tokens": 234781077.0, "step": 6154 }, { "epoch": 0.7829792647245898, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.907815933227539, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8468407392501831, "num_tokens": 234809374.0, "step": 6155 }, { "epoch": 0.7831064750031802, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.897929191589355, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8509635925292969, "num_tokens": 234845585.0, "step": 6156 }, { "epoch": 0.7832336852817707, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.873802185058594, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8608809113502502, "num_tokens": 234883891.0, "step": 6157 }, { "epoch": 0.7833608955603613, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.887811660766602, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8613185882568359, "num_tokens": 234919445.0, "step": 6158 }, { "epoch": 0.7834881058389518, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.825773239135742, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8644746541976929, "num_tokens": 234951465.0, "step": 6159 }, { "epoch": 0.7836153161175423, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.867250442504883, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8549516201019287, "num_tokens": 234989970.0, "step": 6160 }, { "epoch": 0.7837425263961328, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.865577697753906, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8635124564170837, "num_tokens": 235029673.0, "step": 6161 }, { "epoch": 0.7838697366747234, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.838468551635742, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8610746264457703, "num_tokens": 235065389.0, "step": 6162 }, { "epoch": 0.7839969469533138, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.891410827636719, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8738207221031189, "num_tokens": 235099795.0, "step": 6163 }, { "epoch": 0.7841241572319043, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.856595039367676, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8626895546913147, "num_tokens": 235137475.0, "step": 6164 }, { "epoch": 0.7842513675104948, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.857661247253418, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.857142448425293, "num_tokens": 235178454.0, "step": 6165 }, { "epoch": 0.7843785777890854, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.857039451599121, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8803684711456299, "num_tokens": 235216139.0, "step": 6166 }, { "epoch": 0.7845057880676759, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.80305004119873, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8771968483924866, "num_tokens": 235253571.0, "step": 6167 }, { "epoch": 0.7846329983462664, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.862380027770996, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.8459165692329407, "num_tokens": 235293950.0, "step": 6168 }, { "epoch": 0.7847602086248568, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.933196067810059, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8662678003311157, "num_tokens": 235336419.0, "step": 6169 }, { "epoch": 0.7848874189034474, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.81497859954834, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8673215508460999, "num_tokens": 235376039.0, "step": 6170 }, { "epoch": 0.7850146291820379, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.923047065734863, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8505091071128845, "num_tokens": 235412736.0, "step": 6171 }, { "epoch": 0.7851418394606284, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.860808372497559, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8627443909645081, "num_tokens": 235457318.0, "step": 6172 }, { "epoch": 0.785269049739219, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.88025188446045, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8524926900863647, "num_tokens": 235491471.0, "step": 6173 }, { "epoch": 0.7853962600178095, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.845489501953125, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8720072507858276, "num_tokens": 235531324.0, "step": 6174 }, { "epoch": 0.7855234702963999, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.883426666259766, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8496814370155334, "num_tokens": 235573233.0, "step": 6175 }, { "epoch": 0.7856506805749904, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.821670532226562, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8631100654602051, "num_tokens": 235614400.0, "step": 6176 }, { "epoch": 0.785777890853581, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.857527732849121, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8685981035232544, "num_tokens": 235652247.0, "step": 6177 }, { "epoch": 0.7859051011321715, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.892189979553223, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8589239120483398, "num_tokens": 235681810.0, "step": 6178 }, { "epoch": 0.786032311410762, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.805777549743652, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8714147806167603, "num_tokens": 235718422.0, "step": 6179 }, { "epoch": 0.7861595216893525, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.858926773071289, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8579604029655457, "num_tokens": 235758886.0, "step": 6180 }, { "epoch": 0.786286731967943, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.899527549743652, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8660579919815063, "num_tokens": 235792270.0, "step": 6181 }, { "epoch": 0.7864139422465335, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.808809280395508, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8521552681922913, "num_tokens": 235826678.0, "step": 6182 }, { "epoch": 0.786541152525124, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.891979217529297, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8741456270217896, "num_tokens": 235860437.0, "step": 6183 }, { "epoch": 0.7866683628037145, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.878271102905273, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8674405813217163, "num_tokens": 235900865.0, "step": 6184 }, { "epoch": 0.7867955730823051, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.846878051757812, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8479110598564148, "num_tokens": 235937419.0, "step": 6185 }, { "epoch": 0.7869227833608956, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.864191055297852, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8527522087097168, "num_tokens": 235976425.0, "step": 6186 }, { "epoch": 0.787049993639486, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.905233383178711, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8560082316398621, "num_tokens": 236012968.0, "step": 6187 }, { "epoch": 0.7871772039180766, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.827022552490234, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8640444278717041, "num_tokens": 236052462.0, "step": 6188 }, { "epoch": 0.7873044141966671, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.931171417236328, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8778598308563232, "num_tokens": 236087604.0, "step": 6189 }, { "epoch": 0.7874316244752576, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.899564743041992, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8615521192550659, "num_tokens": 236126344.0, "step": 6190 }, { "epoch": 0.7875588347538481, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.842432022094727, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8700903058052063, "num_tokens": 236161913.0, "step": 6191 }, { "epoch": 0.7876860450324387, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.876944541931152, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8660210371017456, "num_tokens": 236200612.0, "step": 6192 }, { "epoch": 0.7878132553110291, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.869077682495117, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8685390949249268, "num_tokens": 236238136.0, "step": 6193 }, { "epoch": 0.7879404655896196, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.96801471710205, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8585410118103027, "num_tokens": 236281303.0, "step": 6194 }, { "epoch": 0.7880676758682101, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.880681037902832, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8591228723526001, "num_tokens": 236322355.0, "step": 6195 }, { "epoch": 0.7881948861468007, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.938319206237793, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8614025712013245, "num_tokens": 236354263.0, "step": 6196 }, { "epoch": 0.7883220964253912, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.928156852722168, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8636201024055481, "num_tokens": 236391581.0, "step": 6197 }, { "epoch": 0.7884493067039817, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.856742858886719, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8713098168373108, "num_tokens": 236426003.0, "step": 6198 }, { "epoch": 0.7885765169825721, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.952445030212402, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8724392652511597, "num_tokens": 236464874.0, "step": 6199 }, { "epoch": 0.7887037272611627, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.905085563659668, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8730930089950562, "num_tokens": 236499826.0, "step": 6200 }, { "epoch": 0.7888309375397532, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.878927230834961, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8743380904197693, "num_tokens": 236538615.0, "step": 6201 }, { "epoch": 0.7889581478183437, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.884987831115723, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8603063821792603, "num_tokens": 236577860.0, "step": 6202 }, { "epoch": 0.7890853580969343, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.853066444396973, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8773370981216431, "num_tokens": 236619041.0, "step": 6203 }, { "epoch": 0.7892125683755248, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.875584602355957, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8656385540962219, "num_tokens": 236655788.0, "step": 6204 }, { "epoch": 0.7893397786541152, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.924101829528809, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8740046620368958, "num_tokens": 236691532.0, "step": 6205 }, { "epoch": 0.7894669889327057, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.825820922851562, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8697024583816528, "num_tokens": 236733792.0, "step": 6206 }, { "epoch": 0.7895941992112963, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.859640121459961, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8751234412193298, "num_tokens": 236775454.0, "step": 6207 }, { "epoch": 0.7897214094898868, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.852644920349121, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8733538389205933, "num_tokens": 236811572.0, "step": 6208 }, { "epoch": 0.7898486197684773, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.861162185668945, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8475114107131958, "num_tokens": 236851018.0, "step": 6209 }, { "epoch": 0.7899758300470678, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.90562915802002, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8678023815155029, "num_tokens": 236891871.0, "step": 6210 }, { "epoch": 0.7901030403256584, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.864913940429688, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8661911487579346, "num_tokens": 236926771.0, "step": 6211 }, { "epoch": 0.7902302506042488, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.914240837097168, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.850365161895752, "num_tokens": 236964768.0, "step": 6212 }, { "epoch": 0.7903574608828393, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.85201358795166, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8516699075698853, "num_tokens": 237005075.0, "step": 6213 }, { "epoch": 0.7904846711614298, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.916895866394043, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8647177219390869, "num_tokens": 237035517.0, "step": 6214 }, { "epoch": 0.7906118814400204, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.856002807617188, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8805674910545349, "num_tokens": 237076319.0, "step": 6215 }, { "epoch": 0.7907390917186109, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.833866119384766, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8538621068000793, "num_tokens": 237117935.0, "step": 6216 }, { "epoch": 0.7908663019972014, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.824953079223633, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.872782826423645, "num_tokens": 237155789.0, "step": 6217 }, { "epoch": 0.7909935122757918, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.898454666137695, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8689591884613037, "num_tokens": 237198870.0, "step": 6218 }, { "epoch": 0.7911207225543824, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.864361763000488, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8676599860191345, "num_tokens": 237243084.0, "step": 6219 }, { "epoch": 0.7912479328329729, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.922015190124512, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8563486933708191, "num_tokens": 237277245.0, "step": 6220 }, { "epoch": 0.7913751431115634, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.84812068939209, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8673579692840576, "num_tokens": 237317841.0, "step": 6221 }, { "epoch": 0.791502353390154, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.962581634521484, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.852908730506897, "num_tokens": 237349327.0, "step": 6222 }, { "epoch": 0.7916295636687445, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.85956859588623, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8487924337387085, "num_tokens": 237396179.0, "step": 6223 }, { "epoch": 0.7917567739473349, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.85770320892334, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8592022061347961, "num_tokens": 237437766.0, "step": 6224 }, { "epoch": 0.7918839842259254, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.858633041381836, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8678497672080994, "num_tokens": 237479239.0, "step": 6225 }, { "epoch": 0.792011194504516, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.856685638427734, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8766207098960876, "num_tokens": 237514641.0, "step": 6226 }, { "epoch": 0.7921384047831065, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.846229553222656, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8736262917518616, "num_tokens": 237555319.0, "step": 6227 }, { "epoch": 0.792265615061697, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.846883773803711, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8653832077980042, "num_tokens": 237597786.0, "step": 6228 }, { "epoch": 0.7923928253402875, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.904720306396484, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8748331069946289, "num_tokens": 237636584.0, "step": 6229 }, { "epoch": 0.792520035618878, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.845170974731445, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.855356752872467, "num_tokens": 237676000.0, "step": 6230 }, { "epoch": 0.7926472458974685, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.906951904296875, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.889541745185852, "num_tokens": 237713003.0, "step": 6231 }, { "epoch": 0.792774456176059, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.908428192138672, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.865716814994812, "num_tokens": 237742054.0, "step": 6232 }, { "epoch": 0.7929016664546495, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.911130905151367, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8561457395553589, "num_tokens": 237783147.0, "step": 6233 }, { "epoch": 0.7930288767332401, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.890643119812012, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8578231334686279, "num_tokens": 237814322.0, "step": 6234 }, { "epoch": 0.7931560870118306, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.877394676208496, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8687107563018799, "num_tokens": 237849278.0, "step": 6235 }, { "epoch": 0.793283297290421, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.84244441986084, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8676340579986572, "num_tokens": 237893434.0, "step": 6236 }, { "epoch": 0.7934105075690115, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.860785484313965, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8714404106140137, "num_tokens": 237932135.0, "step": 6237 }, { "epoch": 0.7935377178476021, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 11.007861137390137, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8578594923019409, "num_tokens": 237966320.0, "step": 6238 }, { "epoch": 0.7936649281261926, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.85602855682373, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8681231737136841, "num_tokens": 238006365.0, "step": 6239 }, { "epoch": 0.7937921384047831, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.92276668548584, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8624323010444641, "num_tokens": 238039307.0, "step": 6240 }, { "epoch": 0.7939193486833737, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.926138877868652, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8497132062911987, "num_tokens": 238078015.0, "step": 6241 }, { "epoch": 0.7940465589619641, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.925756454467773, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8606019020080566, "num_tokens": 238116790.0, "step": 6242 }, { "epoch": 0.7941737692405546, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.899091720581055, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8699309825897217, "num_tokens": 238155330.0, "step": 6243 }, { "epoch": 0.7943009795191451, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.821470260620117, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8481757640838623, "num_tokens": 238194492.0, "step": 6244 }, { "epoch": 0.7944281897977357, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.937965393066406, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8636770844459534, "num_tokens": 238230451.0, "step": 6245 }, { "epoch": 0.7945554000763262, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.870630264282227, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8708164095878601, "num_tokens": 238266331.0, "step": 6246 }, { "epoch": 0.7946826103549167, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.916667938232422, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8529145121574402, "num_tokens": 238307753.0, "step": 6247 }, { "epoch": 0.7948098206335071, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.859864234924316, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8642475605010986, "num_tokens": 238349644.0, "step": 6248 }, { "epoch": 0.7949370309120977, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.942397117614746, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8490225076675415, "num_tokens": 238386308.0, "step": 6249 }, { "epoch": 0.7950642411906882, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.8607177734375, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8479863405227661, "num_tokens": 238427935.0, "step": 6250 }, { "epoch": 0.7951914514692787, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.890022277832031, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8683883547782898, "num_tokens": 238463639.0, "step": 6251 }, { "epoch": 0.7953186617478692, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.948786735534668, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8764596581459045, "num_tokens": 238498295.0, "step": 6252 }, { "epoch": 0.7954458720264598, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.827260971069336, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8630938529968262, "num_tokens": 238538109.0, "step": 6253 }, { "epoch": 0.7955730823050502, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.918087005615234, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8498953580856323, "num_tokens": 238574415.0, "step": 6254 }, { "epoch": 0.7957002925836407, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.835116386413574, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8654484152793884, "num_tokens": 238616368.0, "step": 6255 }, { "epoch": 0.7958275028622313, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.96198844909668, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8678505420684814, "num_tokens": 238650634.0, "step": 6256 }, { "epoch": 0.7959547131408218, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.879569053649902, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.858412504196167, "num_tokens": 238684383.0, "step": 6257 }, { "epoch": 0.7960819234194123, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.877212524414062, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8608938455581665, "num_tokens": 238722863.0, "step": 6258 }, { "epoch": 0.7962091336980028, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.834714889526367, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8642899394035339, "num_tokens": 238765027.0, "step": 6259 }, { "epoch": 0.7963363439765934, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.870351791381836, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8673696517944336, "num_tokens": 238801950.0, "step": 6260 }, { "epoch": 0.7964635542551838, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.87983226776123, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8537992238998413, "num_tokens": 238840963.0, "step": 6261 }, { "epoch": 0.7965907645337743, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.88854694366455, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8499423861503601, "num_tokens": 238876167.0, "step": 6262 }, { "epoch": 0.7967179748123648, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.919095039367676, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8696300983428955, "num_tokens": 238910293.0, "step": 6263 }, { "epoch": 0.7968451850909554, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.905423164367676, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8598915934562683, "num_tokens": 238947279.0, "step": 6264 }, { "epoch": 0.7969723953695459, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.913627624511719, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8521749973297119, "num_tokens": 238984148.0, "step": 6265 }, { "epoch": 0.7970996056481364, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.880431175231934, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8499212265014648, "num_tokens": 239022370.0, "step": 6266 }, { "epoch": 0.7972268159267268, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.911195755004883, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8591115474700928, "num_tokens": 239064542.0, "step": 6267 }, { "epoch": 0.7973540262053174, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.906442642211914, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8615726232528687, "num_tokens": 239103193.0, "step": 6268 }, { "epoch": 0.7974812364839079, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.984536170959473, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8704895973205566, "num_tokens": 239137952.0, "step": 6269 }, { "epoch": 0.7976084467624984, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.916160583496094, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8534170985221863, "num_tokens": 239174600.0, "step": 6270 }, { "epoch": 0.797735657041089, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.933586120605469, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8674123883247375, "num_tokens": 239206157.0, "step": 6271 }, { "epoch": 0.7978628673196795, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.927573204040527, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8677833676338196, "num_tokens": 239243094.0, "step": 6272 }, { "epoch": 0.7979900775982699, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.852485656738281, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8643306493759155, "num_tokens": 239281208.0, "step": 6273 }, { "epoch": 0.7981172878768604, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.993860244750977, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8588426113128662, "num_tokens": 239320159.0, "step": 6274 }, { "epoch": 0.798244498155451, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.88796615600586, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8514390587806702, "num_tokens": 239360160.0, "step": 6275 }, { "epoch": 0.7983717084340415, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.871124267578125, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8650177717208862, "num_tokens": 239398101.0, "step": 6276 }, { "epoch": 0.798498918712632, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.95079231262207, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8784406185150146, "num_tokens": 239438560.0, "step": 6277 }, { "epoch": 0.7986261289912225, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.86351490020752, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8573774099349976, "num_tokens": 239475851.0, "step": 6278 }, { "epoch": 0.798753339269813, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.903496742248535, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8679890632629395, "num_tokens": 239519165.0, "step": 6279 }, { "epoch": 0.7988805495484035, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.938887596130371, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8501816987991333, "num_tokens": 239560602.0, "step": 6280 }, { "epoch": 0.799007759826994, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.993866920471191, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8644236326217651, "num_tokens": 239591933.0, "step": 6281 }, { "epoch": 0.7991349701055845, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.935567855834961, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8571127653121948, "num_tokens": 239625495.0, "step": 6282 }, { "epoch": 0.7992621803841751, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.923587799072266, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8773771524429321, "num_tokens": 239658249.0, "step": 6283 }, { "epoch": 0.7993893906627656, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.953248023986816, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.854960024356842, "num_tokens": 239690194.0, "step": 6284 }, { "epoch": 0.799516600941356, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.915117263793945, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8611149787902832, "num_tokens": 239728315.0, "step": 6285 }, { "epoch": 0.7996438112199465, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.9013671875, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8682105541229248, "num_tokens": 239764505.0, "step": 6286 }, { "epoch": 0.7997710214985371, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.963541984558105, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8473743200302124, "num_tokens": 239809461.0, "step": 6287 }, { "epoch": 0.7998982317771276, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.923739433288574, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8571884036064148, "num_tokens": 239845232.0, "step": 6288 }, { "epoch": 0.8000254420557181, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.934735298156738, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.868565022945404, "num_tokens": 239885489.0, "step": 6289 }, { "epoch": 0.8001526523343087, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.882044792175293, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.85052490234375, "num_tokens": 239921159.0, "step": 6290 }, { "epoch": 0.8002798626128991, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.881437301635742, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8670831918716431, "num_tokens": 239960879.0, "step": 6291 }, { "epoch": 0.8004070728914896, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 11.005393981933594, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8529314994812012, "num_tokens": 239994302.0, "step": 6292 }, { "epoch": 0.8005342831700801, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.831964492797852, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8692561388015747, "num_tokens": 240027759.0, "step": 6293 }, { "epoch": 0.8006614934486707, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.99254322052002, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8582189083099365, "num_tokens": 240065049.0, "step": 6294 }, { "epoch": 0.8007887037272612, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.922365188598633, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8489455580711365, "num_tokens": 240101453.0, "step": 6295 }, { "epoch": 0.8009159140058517, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.932753562927246, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8422197103500366, "num_tokens": 240136831.0, "step": 6296 }, { "epoch": 0.8010431242844421, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.93329906463623, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8615905046463013, "num_tokens": 240173364.0, "step": 6297 }, { "epoch": 0.8011703345630327, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.984155654907227, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8408844470977783, "num_tokens": 240206718.0, "step": 6298 }, { "epoch": 0.8012975448416232, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.923163414001465, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8502050638198853, "num_tokens": 240245633.0, "step": 6299 }, { "epoch": 0.8014247551202137, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.930869102478027, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8762973546981812, "num_tokens": 240283887.0, "step": 6300 }, { "epoch": 0.8015519653988042, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.949384689331055, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8542917966842651, "num_tokens": 240322208.0, "step": 6301 }, { "epoch": 0.8016791756773948, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.963489532470703, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8657256960868835, "num_tokens": 240356154.0, "step": 6302 }, { "epoch": 0.8018063859559852, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.991009712219238, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8607302904129028, "num_tokens": 240389103.0, "step": 6303 }, { "epoch": 0.8019335962345757, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.898242950439453, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8789718747138977, "num_tokens": 240429545.0, "step": 6304 }, { "epoch": 0.8020608065131662, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.955753326416016, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8793325424194336, "num_tokens": 240468556.0, "step": 6305 }, { "epoch": 0.8021880167917568, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.903714179992676, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8637130260467529, "num_tokens": 240508534.0, "step": 6306 }, { "epoch": 0.8023152270703473, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.943105697631836, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8624937534332275, "num_tokens": 240540833.0, "step": 6307 }, { "epoch": 0.8024424373489378, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.95217227935791, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8547117114067078, "num_tokens": 240577768.0, "step": 6308 }, { "epoch": 0.8025696476275284, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.893410682678223, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.858575701713562, "num_tokens": 240610607.0, "step": 6309 }, { "epoch": 0.8026968579061188, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.945483207702637, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8593413829803467, "num_tokens": 240646056.0, "step": 6310 }, { "epoch": 0.8028240681847093, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.948349952697754, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8700170516967773, "num_tokens": 240682513.0, "step": 6311 }, { "epoch": 0.8029512784632998, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.942939758300781, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8635631203651428, "num_tokens": 240716311.0, "step": 6312 }, { "epoch": 0.8030784887418904, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.941550254821777, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8543745279312134, "num_tokens": 240758873.0, "step": 6313 }, { "epoch": 0.8032056990204809, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.931480407714844, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8675143718719482, "num_tokens": 240797163.0, "step": 6314 }, { "epoch": 0.8033329092990714, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.952950477600098, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8739354610443115, "num_tokens": 240835719.0, "step": 6315 }, { "epoch": 0.8034601195776618, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.951306343078613, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8555524349212646, "num_tokens": 240874359.0, "step": 6316 }, { "epoch": 0.8035873298562524, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.936600685119629, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8753295540809631, "num_tokens": 240909125.0, "step": 6317 }, { "epoch": 0.8037145401348429, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.907598495483398, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8666910529136658, "num_tokens": 240949391.0, "step": 6318 }, { "epoch": 0.8038417504134334, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.975519180297852, "learning_rate": 1e-06, "loss": 0.4693, "mean_token_accuracy": 0.8526241183280945, "num_tokens": 240987390.0, "step": 6319 }, { "epoch": 0.803968960692024, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.920225143432617, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8553067445755005, "num_tokens": 241025359.0, "step": 6320 }, { "epoch": 0.8040961709706145, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.890449523925781, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8578850030899048, "num_tokens": 241060532.0, "step": 6321 }, { "epoch": 0.8042233812492049, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.927215576171875, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8692182302474976, "num_tokens": 241098547.0, "step": 6322 }, { "epoch": 0.8043505915277954, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.895520210266113, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.862245500087738, "num_tokens": 241131091.0, "step": 6323 }, { "epoch": 0.804477801806386, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.99905014038086, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8525862693786621, "num_tokens": 241164495.0, "step": 6324 }, { "epoch": 0.8046050120849765, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.902710914611816, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8564735651016235, "num_tokens": 241203777.0, "step": 6325 }, { "epoch": 0.804732222363567, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.881406784057617, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8558565378189087, "num_tokens": 241243219.0, "step": 6326 }, { "epoch": 0.8048594326421575, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.919920921325684, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.8472685813903809, "num_tokens": 241284547.0, "step": 6327 }, { "epoch": 0.804986642920748, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.964324951171875, "learning_rate": 1e-06, "loss": 0.4944, "mean_token_accuracy": 0.8448114395141602, "num_tokens": 241331110.0, "step": 6328 }, { "epoch": 0.8051138531993385, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.920902252197266, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8615233302116394, "num_tokens": 241372720.0, "step": 6329 }, { "epoch": 0.805241063477929, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.925949096679688, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8635285496711731, "num_tokens": 241408173.0, "step": 6330 }, { "epoch": 0.8053682737565195, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.966740608215332, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8687875270843506, "num_tokens": 241445021.0, "step": 6331 }, { "epoch": 0.8054954840351101, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.930656433105469, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8667201995849609, "num_tokens": 241481874.0, "step": 6332 }, { "epoch": 0.8056226943137006, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.924391746520996, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8472098112106323, "num_tokens": 241523140.0, "step": 6333 }, { "epoch": 0.805749904592291, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.902087211608887, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8442466259002686, "num_tokens": 241563424.0, "step": 6334 }, { "epoch": 0.8058771148708815, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.918487548828125, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8748655319213867, "num_tokens": 241600004.0, "step": 6335 }, { "epoch": 0.8060043251494721, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.902819633483887, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8605777025222778, "num_tokens": 241642704.0, "step": 6336 }, { "epoch": 0.8061315354280626, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.860840797424316, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8670390248298645, "num_tokens": 241679176.0, "step": 6337 }, { "epoch": 0.8062587457066531, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.909453392028809, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8668835759162903, "num_tokens": 241721168.0, "step": 6338 }, { "epoch": 0.8063859559852437, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.892533302307129, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8747522234916687, "num_tokens": 241759735.0, "step": 6339 }, { "epoch": 0.8065131662638341, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.957791328430176, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.862621009349823, "num_tokens": 241799445.0, "step": 6340 }, { "epoch": 0.8066403765424246, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.878715515136719, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8663007616996765, "num_tokens": 241839333.0, "step": 6341 }, { "epoch": 0.8067675868210151, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.910252571105957, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8502987623214722, "num_tokens": 241880248.0, "step": 6342 }, { "epoch": 0.8068947970996057, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.89577865600586, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8754516839981079, "num_tokens": 241926717.0, "step": 6343 }, { "epoch": 0.8070220073781962, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.953024864196777, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8530153036117554, "num_tokens": 241966080.0, "step": 6344 }, { "epoch": 0.8071492176567867, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.938668251037598, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8597978949546814, "num_tokens": 242011975.0, "step": 6345 }, { "epoch": 0.8072764279353771, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.914177894592285, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8748117089271545, "num_tokens": 242046320.0, "step": 6346 }, { "epoch": 0.8074036382139677, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.92534065246582, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.863040030002594, "num_tokens": 242085157.0, "step": 6347 }, { "epoch": 0.8075308484925582, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.92912769317627, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8583041429519653, "num_tokens": 242116548.0, "step": 6348 }, { "epoch": 0.8076580587711487, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.97669792175293, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.868466854095459, "num_tokens": 242151883.0, "step": 6349 }, { "epoch": 0.8077852690497392, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.931819915771484, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8592543601989746, "num_tokens": 242190486.0, "step": 6350 }, { "epoch": 0.8079124793283298, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.946029663085938, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8659422397613525, "num_tokens": 242232720.0, "step": 6351 }, { "epoch": 0.8080396896069202, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.854183197021484, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8636656403541565, "num_tokens": 242273926.0, "step": 6352 }, { "epoch": 0.8081668998855107, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.963611602783203, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8519691228866577, "num_tokens": 242313484.0, "step": 6353 }, { "epoch": 0.8082941101641012, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.977917671203613, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8558290004730225, "num_tokens": 242352101.0, "step": 6354 }, { "epoch": 0.8084213204426918, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.89151668548584, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8774599432945251, "num_tokens": 242392958.0, "step": 6355 }, { "epoch": 0.8085485307212823, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.861990928649902, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8635693788528442, "num_tokens": 242432327.0, "step": 6356 }, { "epoch": 0.8086757409998728, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.97834300994873, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.8418506383895874, "num_tokens": 242473714.0, "step": 6357 }, { "epoch": 0.8088029512784632, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.889930725097656, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8695039749145508, "num_tokens": 242513179.0, "step": 6358 }, { "epoch": 0.8089301615570538, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.973518371582031, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8632415533065796, "num_tokens": 242549398.0, "step": 6359 }, { "epoch": 0.8090573718356443, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.906755447387695, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8573894500732422, "num_tokens": 242583705.0, "step": 6360 }, { "epoch": 0.8091845821142348, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.887539863586426, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8666778206825256, "num_tokens": 242625900.0, "step": 6361 }, { "epoch": 0.8093117923928254, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.92651653289795, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8515949249267578, "num_tokens": 242665766.0, "step": 6362 }, { "epoch": 0.8094390026714159, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.908695220947266, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8653910160064697, "num_tokens": 242707634.0, "step": 6363 }, { "epoch": 0.8095662129500064, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.887038230895996, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8577800393104553, "num_tokens": 242746791.0, "step": 6364 }, { "epoch": 0.8096934232285968, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.918391227722168, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8758194446563721, "num_tokens": 242784861.0, "step": 6365 }, { "epoch": 0.8098206335071874, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.925823211669922, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8773012161254883, "num_tokens": 242817612.0, "step": 6366 }, { "epoch": 0.8099478437857779, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.911023139953613, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8692094683647156, "num_tokens": 242854951.0, "step": 6367 }, { "epoch": 0.8100750540643684, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.858633995056152, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8771355152130127, "num_tokens": 242894166.0, "step": 6368 }, { "epoch": 0.8102022643429589, "ewc_loss": 0.021484375, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.93272590637207, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8678339123725891, "num_tokens": 242936230.0, "step": 6369 }, { "epoch": 0.8103294746215495, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.967838287353516, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8727261424064636, "num_tokens": 242973148.0, "step": 6370 }, { "epoch": 0.8104566849001399, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.9916353225708, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8636137247085571, "num_tokens": 243013029.0, "step": 6371 }, { "epoch": 0.8105838951787304, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.862622261047363, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.861138641834259, "num_tokens": 243054847.0, "step": 6372 }, { "epoch": 0.810711105457321, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.953780174255371, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8563061952590942, "num_tokens": 243096524.0, "step": 6373 }, { "epoch": 0.8108383157359115, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.952703475952148, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.854935348033905, "num_tokens": 243132397.0, "step": 6374 }, { "epoch": 0.810965526014502, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.890778541564941, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8493001461029053, "num_tokens": 243171790.0, "step": 6375 }, { "epoch": 0.8110927362930925, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.967235565185547, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8617494702339172, "num_tokens": 243208217.0, "step": 6376 }, { "epoch": 0.811219946571683, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.952472686767578, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8668115139007568, "num_tokens": 243245260.0, "step": 6377 }, { "epoch": 0.8113471568502735, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.988304138183594, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8705930113792419, "num_tokens": 243283924.0, "step": 6378 }, { "epoch": 0.811474367128864, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.881178855895996, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8706154823303223, "num_tokens": 243321977.0, "step": 6379 }, { "epoch": 0.8116015774074545, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.938650131225586, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8588485717773438, "num_tokens": 243359349.0, "step": 6380 }, { "epoch": 0.8117287876860451, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.983317375183105, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8660064935684204, "num_tokens": 243395722.0, "step": 6381 }, { "epoch": 0.8118559979646356, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.91685962677002, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8518770337104797, "num_tokens": 243429022.0, "step": 6382 }, { "epoch": 0.811983208243226, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.977787017822266, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.8414818048477173, "num_tokens": 243468474.0, "step": 6383 }, { "epoch": 0.8121104185218165, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.910503387451172, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.878073513507843, "num_tokens": 243503045.0, "step": 6384 }, { "epoch": 0.8122376288004071, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.005412101745605, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8520689010620117, "num_tokens": 243543865.0, "step": 6385 }, { "epoch": 0.8123648390789976, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.891172409057617, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8559325933456421, "num_tokens": 243581258.0, "step": 6386 }, { "epoch": 0.8124920493575881, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.901604652404785, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.856706440448761, "num_tokens": 243622777.0, "step": 6387 }, { "epoch": 0.8126192596361786, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.963034629821777, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8499808311462402, "num_tokens": 243664549.0, "step": 6388 }, { "epoch": 0.8127464699147691, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.907439231872559, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8561455011367798, "num_tokens": 243710408.0, "step": 6389 }, { "epoch": 0.8128736801933596, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.908405303955078, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.863518238067627, "num_tokens": 243752598.0, "step": 6390 }, { "epoch": 0.8130008904719501, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.981175422668457, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8592863082885742, "num_tokens": 243789805.0, "step": 6391 }, { "epoch": 0.8131281007505406, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.938150405883789, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.872681736946106, "num_tokens": 243824124.0, "step": 6392 }, { "epoch": 0.8132553110291312, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.000603675842285, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8621318936347961, "num_tokens": 243863448.0, "step": 6393 }, { "epoch": 0.8133825213077217, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.934197425842285, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8654720783233643, "num_tokens": 243902085.0, "step": 6394 }, { "epoch": 0.8135097315863121, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.991984367370605, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8571573495864868, "num_tokens": 243945349.0, "step": 6395 }, { "epoch": 0.8136369418649027, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.991964340209961, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8605570793151855, "num_tokens": 243982221.0, "step": 6396 }, { "epoch": 0.8137641521434932, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.957903861999512, "learning_rate": 1e-06, "loss": 0.4649, "mean_token_accuracy": 0.8488436937332153, "num_tokens": 244012576.0, "step": 6397 }, { "epoch": 0.8138913624220837, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.979949951171875, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8694429993629456, "num_tokens": 244052089.0, "step": 6398 }, { "epoch": 0.8140185727006742, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.920680046081543, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8634130954742432, "num_tokens": 244088712.0, "step": 6399 }, { "epoch": 0.8141457829792648, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.046710968017578, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8632941842079163, "num_tokens": 244127655.0, "step": 6400 }, { "epoch": 0.8142729932578552, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.961955070495605, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8677897453308105, "num_tokens": 244158588.0, "step": 6401 }, { "epoch": 0.8144002035364457, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.971406936645508, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8810583353042603, "num_tokens": 244193872.0, "step": 6402 }, { "epoch": 0.8145274138150362, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.912755966186523, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.8472940325737, "num_tokens": 244234055.0, "step": 6403 }, { "epoch": 0.8146546240936268, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.9233980178833, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8640168905258179, "num_tokens": 244275163.0, "step": 6404 }, { "epoch": 0.8147818343722173, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.910719871520996, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8597877025604248, "num_tokens": 244315241.0, "step": 6405 }, { "epoch": 0.8149090446508078, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.974212646484375, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8773128986358643, "num_tokens": 244351947.0, "step": 6406 }, { "epoch": 0.8150362549293982, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.946117401123047, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8595836162567139, "num_tokens": 244387706.0, "step": 6407 }, { "epoch": 0.8151634652079888, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.931791305541992, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8707918524742126, "num_tokens": 244428447.0, "step": 6408 }, { "epoch": 0.8152906754865793, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.943781852722168, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8478379249572754, "num_tokens": 244471191.0, "step": 6409 }, { "epoch": 0.8154178857651698, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.004377365112305, "learning_rate": 1e-06, "loss": 0.4989, "mean_token_accuracy": 0.8442595601081848, "num_tokens": 244509198.0, "step": 6410 }, { "epoch": 0.8155450960437604, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.047955513000488, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8546431064605713, "num_tokens": 244549677.0, "step": 6411 }, { "epoch": 0.8156723063223509, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.891809463500977, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8501592874526978, "num_tokens": 244583586.0, "step": 6412 }, { "epoch": 0.8157995166009414, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.973724365234375, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8695245981216431, "num_tokens": 244621220.0, "step": 6413 }, { "epoch": 0.8159267268795318, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.92703628540039, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8691725730895996, "num_tokens": 244655941.0, "step": 6414 }, { "epoch": 0.8160539371581224, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.94221019744873, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8791443109512329, "num_tokens": 244685807.0, "step": 6415 }, { "epoch": 0.8161811474367129, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.965597152709961, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8600696325302124, "num_tokens": 244726866.0, "step": 6416 }, { "epoch": 0.8163083577153034, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84696102142334, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8536723852157593, "num_tokens": 244767273.0, "step": 6417 }, { "epoch": 0.8164355679938939, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.99638843536377, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8580483198165894, "num_tokens": 244806111.0, "step": 6418 }, { "epoch": 0.8165627782724845, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.937969207763672, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8731204271316528, "num_tokens": 244841706.0, "step": 6419 }, { "epoch": 0.8166899885510749, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.902503967285156, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8640214204788208, "num_tokens": 244883908.0, "step": 6420 }, { "epoch": 0.8168171988296654, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.940387725830078, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.854327380657196, "num_tokens": 244919090.0, "step": 6421 }, { "epoch": 0.8169444091082559, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.934245109558105, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8769027590751648, "num_tokens": 244961141.0, "step": 6422 }, { "epoch": 0.8170716193868465, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.949323654174805, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8464999794960022, "num_tokens": 245004503.0, "step": 6423 }, { "epoch": 0.817198829665437, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.943267822265625, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8608864545822144, "num_tokens": 245045042.0, "step": 6424 }, { "epoch": 0.8173260399440275, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.939962387084961, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8605244755744934, "num_tokens": 245085416.0, "step": 6425 }, { "epoch": 0.8174532502226179, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.980206489562988, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8521161079406738, "num_tokens": 245116137.0, "step": 6426 }, { "epoch": 0.8175804605012085, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.918851852416992, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8746247291564941, "num_tokens": 245159181.0, "step": 6427 }, { "epoch": 0.817707670779799, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.950045585632324, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8536790013313293, "num_tokens": 245205199.0, "step": 6428 }, { "epoch": 0.8178348810583895, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.935897827148438, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8674125671386719, "num_tokens": 245242455.0, "step": 6429 }, { "epoch": 0.8179620913369801, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.967605590820312, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8815655708312988, "num_tokens": 245280675.0, "step": 6430 }, { "epoch": 0.8180893016155706, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.938392639160156, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8776295185089111, "num_tokens": 245316947.0, "step": 6431 }, { "epoch": 0.818216511894161, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.988853454589844, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8665599822998047, "num_tokens": 245355145.0, "step": 6432 }, { "epoch": 0.8183437221727515, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.961386680603027, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8732746839523315, "num_tokens": 245390282.0, "step": 6433 }, { "epoch": 0.8184709324513421, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.959092140197754, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8716050386428833, "num_tokens": 245430262.0, "step": 6434 }, { "epoch": 0.8185981427299326, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.021156311035156, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8596847057342529, "num_tokens": 245470874.0, "step": 6435 }, { "epoch": 0.8187253530085231, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.983722686767578, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8691906929016113, "num_tokens": 245506626.0, "step": 6436 }, { "epoch": 0.8188525632871136, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.974489212036133, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.845770001411438, "num_tokens": 245546794.0, "step": 6437 }, { "epoch": 0.8189797735657041, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.978082656860352, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8658631443977356, "num_tokens": 245589544.0, "step": 6438 }, { "epoch": 0.8191069838442946, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.959314346313477, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8788636922836304, "num_tokens": 245628889.0, "step": 6439 }, { "epoch": 0.8192341941228851, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.040599822998047, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8586452007293701, "num_tokens": 245674615.0, "step": 6440 }, { "epoch": 0.8193614044014756, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.964689254760742, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8620994091033936, "num_tokens": 245713966.0, "step": 6441 }, { "epoch": 0.8194886146800662, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.980813026428223, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8668343424797058, "num_tokens": 245750035.0, "step": 6442 }, { "epoch": 0.8196158249586567, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.989311218261719, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8778238892555237, "num_tokens": 245788506.0, "step": 6443 }, { "epoch": 0.8197430352372471, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.923542976379395, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8658850789070129, "num_tokens": 245826547.0, "step": 6444 }, { "epoch": 0.8198702455158376, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.981487274169922, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8530885577201843, "num_tokens": 245874363.0, "step": 6445 }, { "epoch": 0.8199974557944282, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.989791870117188, "learning_rate": 1e-06, "loss": 0.4873, "mean_token_accuracy": 0.8436093330383301, "num_tokens": 245917813.0, "step": 6446 }, { "epoch": 0.8201246660730187, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.954658508300781, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8588588833808899, "num_tokens": 245952177.0, "step": 6447 }, { "epoch": 0.8202518763516092, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.940394401550293, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8647058010101318, "num_tokens": 245992897.0, "step": 6448 }, { "epoch": 0.8203790866301998, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.985433578491211, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8710248470306396, "num_tokens": 246029657.0, "step": 6449 }, { "epoch": 0.8205062969087902, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.981790542602539, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8535361886024475, "num_tokens": 246072425.0, "step": 6450 }, { "epoch": 0.8206335071873807, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.966914176940918, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8803227543830872, "num_tokens": 246113131.0, "step": 6451 }, { "epoch": 0.8207607174659712, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.913385391235352, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8659120798110962, "num_tokens": 246154548.0, "step": 6452 }, { "epoch": 0.8208879277445618, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.007734298706055, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8594435453414917, "num_tokens": 246189518.0, "step": 6453 }, { "epoch": 0.8210151380231523, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.978348731994629, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8581076860427856, "num_tokens": 246235248.0, "step": 6454 }, { "epoch": 0.8211423483017428, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.914020538330078, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8555072546005249, "num_tokens": 246277016.0, "step": 6455 }, { "epoch": 0.8212695585803332, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.969573974609375, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.860369861125946, "num_tokens": 246312352.0, "step": 6456 }, { "epoch": 0.8213967688589238, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.993830680847168, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8489247560501099, "num_tokens": 246355079.0, "step": 6457 }, { "epoch": 0.8215239791375143, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.007413864135742, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8508855104446411, "num_tokens": 246391073.0, "step": 6458 }, { "epoch": 0.8216511894161048, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.965219497680664, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8693711757659912, "num_tokens": 246435407.0, "step": 6459 }, { "epoch": 0.8217783996946953, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.957707405090332, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8494827747344971, "num_tokens": 246475303.0, "step": 6460 }, { "epoch": 0.8219056099732859, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.909771919250488, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8604760766029358, "num_tokens": 246515237.0, "step": 6461 }, { "epoch": 0.8220328202518764, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.021367073059082, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8576973676681519, "num_tokens": 246554958.0, "step": 6462 }, { "epoch": 0.8221600305304668, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.892036437988281, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8627031445503235, "num_tokens": 246592327.0, "step": 6463 }, { "epoch": 0.8222872408090574, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.917288780212402, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8667021989822388, "num_tokens": 246626965.0, "step": 6464 }, { "epoch": 0.8224144510876479, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.98044204711914, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8587751388549805, "num_tokens": 246662049.0, "step": 6465 }, { "epoch": 0.8225416613662384, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.9795560836792, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8641442060470581, "num_tokens": 246696989.0, "step": 6466 }, { "epoch": 0.8226688716448289, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.007190704345703, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8660241365432739, "num_tokens": 246738418.0, "step": 6467 }, { "epoch": 0.8227960819234195, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.959190368652344, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8633078336715698, "num_tokens": 246776779.0, "step": 6468 }, { "epoch": 0.8229232922020099, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.955364227294922, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8686143755912781, "num_tokens": 246814029.0, "step": 6469 }, { "epoch": 0.8230505024806004, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.951693534851074, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8536602258682251, "num_tokens": 246856390.0, "step": 6470 }, { "epoch": 0.8231777127591909, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.988625526428223, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8636823892593384, "num_tokens": 246895563.0, "step": 6471 }, { "epoch": 0.8233049230377815, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.990797996520996, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.859606146812439, "num_tokens": 246935956.0, "step": 6472 }, { "epoch": 0.823432133316372, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.930806159973145, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8587404489517212, "num_tokens": 246984765.0, "step": 6473 }, { "epoch": 0.8235593435949625, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.122076034545898, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8653038740158081, "num_tokens": 247017086.0, "step": 6474 }, { "epoch": 0.8236865538735529, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.940047264099121, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8677393198013306, "num_tokens": 247056195.0, "step": 6475 }, { "epoch": 0.8238137641521435, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.08979320526123, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8758969902992249, "num_tokens": 247094063.0, "step": 6476 }, { "epoch": 0.823940974430734, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.983294486999512, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8611239194869995, "num_tokens": 247131743.0, "step": 6477 }, { "epoch": 0.8240681847093245, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.037252426147461, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8604069352149963, "num_tokens": 247169477.0, "step": 6478 }, { "epoch": 0.824195394987915, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.962010383605957, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8716265559196472, "num_tokens": 247209159.0, "step": 6479 }, { "epoch": 0.8243226052665056, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.928391456604004, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8642230033874512, "num_tokens": 247247603.0, "step": 6480 }, { "epoch": 0.824449815545096, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.003119468688965, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8726255297660828, "num_tokens": 247290184.0, "step": 6481 }, { "epoch": 0.8245770258236865, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.976198196411133, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8449794054031372, "num_tokens": 247331862.0, "step": 6482 }, { "epoch": 0.824704236102277, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.947547912597656, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8626422882080078, "num_tokens": 247368271.0, "step": 6483 }, { "epoch": 0.8248314463808676, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.059775352478027, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8616307973861694, "num_tokens": 247408077.0, "step": 6484 }, { "epoch": 0.8249586566594581, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.02278995513916, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8692506551742554, "num_tokens": 247447408.0, "step": 6485 }, { "epoch": 0.8250858669380486, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.954805374145508, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8736366033554077, "num_tokens": 247489863.0, "step": 6486 }, { "epoch": 0.8252130772166391, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.908214569091797, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8588806390762329, "num_tokens": 247532233.0, "step": 6487 }, { "epoch": 0.8253402874952296, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.03150749206543, "learning_rate": 1e-06, "loss": 0.5081, "mean_token_accuracy": 0.8386366367340088, "num_tokens": 247570719.0, "step": 6488 }, { "epoch": 0.8254674977738201, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.978886604309082, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8684769868850708, "num_tokens": 247603302.0, "step": 6489 }, { "epoch": 0.8255947080524106, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.93167495727539, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8654301166534424, "num_tokens": 247642837.0, "step": 6490 }, { "epoch": 0.8257219183310012, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.944933891296387, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8594458699226379, "num_tokens": 247685246.0, "step": 6491 }, { "epoch": 0.8258491286095917, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.02730655670166, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8537026047706604, "num_tokens": 247720871.0, "step": 6492 }, { "epoch": 0.8259763388881821, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.896913528442383, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8636141419410706, "num_tokens": 247759796.0, "step": 6493 }, { "epoch": 0.8261035491667726, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.984626770019531, "learning_rate": 1e-06, "loss": 0.501, "mean_token_accuracy": 0.8416482210159302, "num_tokens": 247806486.0, "step": 6494 }, { "epoch": 0.8262307594453632, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.948742866516113, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8713454604148865, "num_tokens": 247842031.0, "step": 6495 }, { "epoch": 0.8263579697239537, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.911421775817871, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8665865659713745, "num_tokens": 247883438.0, "step": 6496 }, { "epoch": 0.8264851800025442, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.012797355651855, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8706172704696655, "num_tokens": 247921491.0, "step": 6497 }, { "epoch": 0.8266123902811348, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.949799537658691, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8724743723869324, "num_tokens": 247957454.0, "step": 6498 }, { "epoch": 0.8267396005597252, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.984580993652344, "learning_rate": 1e-06, "loss": 0.4819, "mean_token_accuracy": 0.8476487994194031, "num_tokens": 247997705.0, "step": 6499 }, { "epoch": 0.8268668108383157, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.982766151428223, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8581897616386414, "num_tokens": 248034536.0, "step": 6500 }, { "epoch": 0.8269940211169062, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.967508316040039, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8637632131576538, "num_tokens": 248080916.0, "step": 6501 }, { "epoch": 0.8271212313954968, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.995357513427734, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.861531674861908, "num_tokens": 248119719.0, "step": 6502 }, { "epoch": 0.8272484416740873, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.96141242980957, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8598138689994812, "num_tokens": 248156221.0, "step": 6503 }, { "epoch": 0.8273756519526778, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.017278671264648, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8668317198753357, "num_tokens": 248191780.0, "step": 6504 }, { "epoch": 0.8275028622312682, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.944488525390625, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8575214147567749, "num_tokens": 248228594.0, "step": 6505 }, { "epoch": 0.8276300725098588, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.001784324645996, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8670322299003601, "num_tokens": 248260372.0, "step": 6506 }, { "epoch": 0.8277572827884493, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.987645149230957, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8543057441711426, "num_tokens": 248299361.0, "step": 6507 }, { "epoch": 0.8278844930670398, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.932303428649902, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8541859984397888, "num_tokens": 248337370.0, "step": 6508 }, { "epoch": 0.8280117033456303, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.988736152648926, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8650754690170288, "num_tokens": 248377541.0, "step": 6509 }, { "epoch": 0.8281389136242209, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.9257230758667, "learning_rate": 1e-06, "loss": 0.4928, "mean_token_accuracy": 0.8451662063598633, "num_tokens": 248413928.0, "step": 6510 }, { "epoch": 0.8282661239028114, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.995420455932617, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8624701499938965, "num_tokens": 248453283.0, "step": 6511 }, { "epoch": 0.8283933341814018, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.961190223693848, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8678713440895081, "num_tokens": 248490286.0, "step": 6512 }, { "epoch": 0.8285205444599923, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.002573013305664, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8478586673736572, "num_tokens": 248526349.0, "step": 6513 }, { "epoch": 0.8286477547385829, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.969494819641113, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8558957576751709, "num_tokens": 248561751.0, "step": 6514 }, { "epoch": 0.8287749650171734, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.967494010925293, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8766798973083496, "num_tokens": 248601921.0, "step": 6515 }, { "epoch": 0.8289021752957639, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.986543655395508, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8738777041435242, "num_tokens": 248641218.0, "step": 6516 }, { "epoch": 0.8290293855743545, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.05778694152832, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8637338876724243, "num_tokens": 248672616.0, "step": 6517 }, { "epoch": 0.8291565958529449, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.991572380065918, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8656381368637085, "num_tokens": 248708212.0, "step": 6518 }, { "epoch": 0.8292838061315354, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.900080680847168, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8677980899810791, "num_tokens": 248746575.0, "step": 6519 }, { "epoch": 0.8294110164101259, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.083503723144531, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8584088087081909, "num_tokens": 248789423.0, "step": 6520 }, { "epoch": 0.8295382266887165, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.996204376220703, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8648315072059631, "num_tokens": 248824044.0, "step": 6521 }, { "epoch": 0.829665436967307, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.993552207946777, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8597177863121033, "num_tokens": 248860805.0, "step": 6522 }, { "epoch": 0.8297926472458975, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.002304077148438, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8684130907058716, "num_tokens": 248898958.0, "step": 6523 }, { "epoch": 0.8299198575244879, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.994333267211914, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8621677160263062, "num_tokens": 248938847.0, "step": 6524 }, { "epoch": 0.8300470678030785, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.98165225982666, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8602213859558105, "num_tokens": 248977944.0, "step": 6525 }, { "epoch": 0.830174278081669, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.998987197875977, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8555357456207275, "num_tokens": 249014610.0, "step": 6526 }, { "epoch": 0.8303014883602595, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.037504196166992, "learning_rate": 1e-06, "loss": 0.525, "mean_token_accuracy": 0.8326929807662964, "num_tokens": 249053869.0, "step": 6527 }, { "epoch": 0.83042869863885, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.995000839233398, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8552786111831665, "num_tokens": 249096874.0, "step": 6528 }, { "epoch": 0.8305559089174406, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.97052001953125, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8618671894073486, "num_tokens": 249133141.0, "step": 6529 }, { "epoch": 0.830683119196031, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.057168006896973, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8675980567932129, "num_tokens": 249163813.0, "step": 6530 }, { "epoch": 0.8308103294746215, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.979424476623535, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8548104763031006, "num_tokens": 249202935.0, "step": 6531 }, { "epoch": 0.830937539753212, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.96538257598877, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8695477247238159, "num_tokens": 249241061.0, "step": 6532 }, { "epoch": 0.8310647500318026, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.083536148071289, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8506249189376831, "num_tokens": 249277863.0, "step": 6533 }, { "epoch": 0.8311919603103931, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.958049774169922, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8620619773864746, "num_tokens": 249318047.0, "step": 6534 }, { "epoch": 0.8313191705889836, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.015896797180176, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8568233847618103, "num_tokens": 249362095.0, "step": 6535 }, { "epoch": 0.831446380867574, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.928601264953613, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8761431574821472, "num_tokens": 249403362.0, "step": 6536 }, { "epoch": 0.8315735911461646, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.999457359313965, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8694804310798645, "num_tokens": 249438051.0, "step": 6537 }, { "epoch": 0.8317008014247551, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.029695510864258, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8671910166740417, "num_tokens": 249477185.0, "step": 6538 }, { "epoch": 0.8318280117033456, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.028767585754395, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8556520938873291, "num_tokens": 249516880.0, "step": 6539 }, { "epoch": 0.8319552219819362, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.97708511352539, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8720809817314148, "num_tokens": 249555864.0, "step": 6540 }, { "epoch": 0.8320824322605267, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.043946266174316, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8616491556167603, "num_tokens": 249590566.0, "step": 6541 }, { "epoch": 0.8322096425391171, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.10568904876709, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8746974468231201, "num_tokens": 249631523.0, "step": 6542 }, { "epoch": 0.8323368528177076, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.025108337402344, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8712636232376099, "num_tokens": 249674113.0, "step": 6543 }, { "epoch": 0.8324640630962982, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.058271408081055, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8726204633712769, "num_tokens": 249712934.0, "step": 6544 }, { "epoch": 0.8325912733748887, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.013115882873535, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8675951957702637, "num_tokens": 249750295.0, "step": 6545 }, { "epoch": 0.8327184836534792, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.059459686279297, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8669310808181763, "num_tokens": 249790289.0, "step": 6546 }, { "epoch": 0.8328456939320698, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.079277992248535, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8761835098266602, "num_tokens": 249822873.0, "step": 6547 }, { "epoch": 0.8329729042106602, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.045948028564453, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8634986877441406, "num_tokens": 249860564.0, "step": 6548 }, { "epoch": 0.8331001144892507, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.029010772705078, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8627565503120422, "num_tokens": 249901448.0, "step": 6549 }, { "epoch": 0.8332273247678412, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.03966236114502, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8528575897216797, "num_tokens": 249933524.0, "step": 6550 }, { "epoch": 0.8333545350464318, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.027567863464355, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8584704995155334, "num_tokens": 249971856.0, "step": 6551 }, { "epoch": 0.8334817453250223, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.034292221069336, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.854953408241272, "num_tokens": 250013940.0, "step": 6552 }, { "epoch": 0.8336089556036128, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.065659523010254, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8463982343673706, "num_tokens": 250044826.0, "step": 6553 }, { "epoch": 0.8337361658822032, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.052348136901855, "learning_rate": 1e-06, "loss": 0.4672, "mean_token_accuracy": 0.8465733528137207, "num_tokens": 250087409.0, "step": 6554 }, { "epoch": 0.8338633761607938, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.940610885620117, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8695774674415588, "num_tokens": 250126236.0, "step": 6555 }, { "epoch": 0.8339905864393843, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.070555686950684, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8706226944923401, "num_tokens": 250163559.0, "step": 6556 }, { "epoch": 0.8341177967179748, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.024560928344727, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8640099763870239, "num_tokens": 250203536.0, "step": 6557 }, { "epoch": 0.8342450069965653, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.045370101928711, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8676869869232178, "num_tokens": 250237455.0, "step": 6558 }, { "epoch": 0.8343722172751559, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.006144523620605, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8716254234313965, "num_tokens": 250274559.0, "step": 6559 }, { "epoch": 0.8344994275537464, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.049181938171387, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8644418120384216, "num_tokens": 250309392.0, "step": 6560 }, { "epoch": 0.8346266378323368, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.983621597290039, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8681341409683228, "num_tokens": 250352010.0, "step": 6561 }, { "epoch": 0.8347538481109273, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.993240356445312, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8508802652359009, "num_tokens": 250385745.0, "step": 6562 }, { "epoch": 0.8348810583895179, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.04297924041748, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8606640100479126, "num_tokens": 250426055.0, "step": 6563 }, { "epoch": 0.8350082686681084, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.951517105102539, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8599756360054016, "num_tokens": 250468795.0, "step": 6564 }, { "epoch": 0.8351354789466989, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.03426456451416, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8614639043807983, "num_tokens": 250509115.0, "step": 6565 }, { "epoch": 0.8352626892252895, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.991048812866211, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8626442551612854, "num_tokens": 250547770.0, "step": 6566 }, { "epoch": 0.8353898995038799, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.982558250427246, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.860349714756012, "num_tokens": 250589737.0, "step": 6567 }, { "epoch": 0.8355171097824704, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.117871284484863, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8706849217414856, "num_tokens": 250623058.0, "step": 6568 }, { "epoch": 0.8356443200610609, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.053842544555664, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8624547719955444, "num_tokens": 250660526.0, "step": 6569 }, { "epoch": 0.8357715303396515, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.947894096374512, "learning_rate": 1e-06, "loss": 0.4871, "mean_token_accuracy": 0.846532940864563, "num_tokens": 250698218.0, "step": 6570 }, { "epoch": 0.835898740618242, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.093838691711426, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8551982641220093, "num_tokens": 250743410.0, "step": 6571 }, { "epoch": 0.8360259508968325, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.989532470703125, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8722195625305176, "num_tokens": 250781919.0, "step": 6572 }, { "epoch": 0.8361531611754229, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.029192924499512, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8602403402328491, "num_tokens": 250817051.0, "step": 6573 }, { "epoch": 0.8362803714540135, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.082817077636719, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8603326082229614, "num_tokens": 250857241.0, "step": 6574 }, { "epoch": 0.836407581732604, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.904953002929688, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8643054962158203, "num_tokens": 250903591.0, "step": 6575 }, { "epoch": 0.8365347920111945, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.009247779846191, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8562241792678833, "num_tokens": 250945788.0, "step": 6576 }, { "epoch": 0.836662002289785, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.05183219909668, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8561081886291504, "num_tokens": 250984884.0, "step": 6577 }, { "epoch": 0.8367892125683756, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.99832820892334, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8746951818466187, "num_tokens": 251024347.0, "step": 6578 }, { "epoch": 0.836916422846966, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.012011528015137, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8840151429176331, "num_tokens": 251068527.0, "step": 6579 }, { "epoch": 0.8370436331255565, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.984779357910156, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8673540949821472, "num_tokens": 251109914.0, "step": 6580 }, { "epoch": 0.837170843404147, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.073119163513184, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8540260791778564, "num_tokens": 251144760.0, "step": 6581 }, { "epoch": 0.8372980536827376, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.056818962097168, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8610355257987976, "num_tokens": 251184456.0, "step": 6582 }, { "epoch": 0.8374252639613281, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.966373443603516, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8551221489906311, "num_tokens": 251229493.0, "step": 6583 }, { "epoch": 0.8375524742399186, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.965950965881348, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8679654598236084, "num_tokens": 251270886.0, "step": 6584 }, { "epoch": 0.837679684518509, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.015545845031738, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.87132728099823, "num_tokens": 251304833.0, "step": 6585 }, { "epoch": 0.8378068947970996, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.981457710266113, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8645554780960083, "num_tokens": 251334196.0, "step": 6586 }, { "epoch": 0.8379341050756901, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.904814720153809, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8521680235862732, "num_tokens": 251380278.0, "step": 6587 }, { "epoch": 0.8380613153542806, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.029458999633789, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8570080399513245, "num_tokens": 251418432.0, "step": 6588 }, { "epoch": 0.8381885256328712, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.929849624633789, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8608306646347046, "num_tokens": 251453747.0, "step": 6589 }, { "epoch": 0.8383157359114617, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.072151184082031, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8627126812934875, "num_tokens": 251492439.0, "step": 6590 }, { "epoch": 0.8384429461900521, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.95157527923584, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.863614022731781, "num_tokens": 251527945.0, "step": 6591 }, { "epoch": 0.8385701564686426, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.015658378601074, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8715258836746216, "num_tokens": 251567161.0, "step": 6592 }, { "epoch": 0.8386973667472332, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.972127914428711, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8623541593551636, "num_tokens": 251601189.0, "step": 6593 }, { "epoch": 0.8388245770258237, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.048425674438477, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8611592650413513, "num_tokens": 251630846.0, "step": 6594 }, { "epoch": 0.8389517873044142, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.919361114501953, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8703689575195312, "num_tokens": 251665701.0, "step": 6595 }, { "epoch": 0.8390789975830047, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.031291961669922, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.855467677116394, "num_tokens": 251699909.0, "step": 6596 }, { "epoch": 0.8392062078615952, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.119232177734375, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8598731756210327, "num_tokens": 251734052.0, "step": 6597 }, { "epoch": 0.8393334181401857, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.938155174255371, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8515213131904602, "num_tokens": 251772518.0, "step": 6598 }, { "epoch": 0.8394606284187762, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.02043342590332, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8608144521713257, "num_tokens": 251816378.0, "step": 6599 }, { "epoch": 0.8395878386973668, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.028030395507812, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8626202344894409, "num_tokens": 251851264.0, "step": 6600 }, { "epoch": 0.8397150489759573, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.056877136230469, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8645226955413818, "num_tokens": 251888906.0, "step": 6601 }, { "epoch": 0.8398422592545478, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.98888111114502, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8687989115715027, "num_tokens": 251930641.0, "step": 6602 }, { "epoch": 0.8399694695331382, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.016485214233398, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8434897661209106, "num_tokens": 251968929.0, "step": 6603 }, { "epoch": 0.8400966798117288, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.100919723510742, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8555001020431519, "num_tokens": 252006751.0, "step": 6604 }, { "epoch": 0.8402238900903193, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.991776466369629, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.858155369758606, "num_tokens": 252045861.0, "step": 6605 }, { "epoch": 0.8403511003689098, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.04133129119873, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8556504249572754, "num_tokens": 252084412.0, "step": 6606 }, { "epoch": 0.8404783106475003, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.98695182800293, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8668787479400635, "num_tokens": 252120576.0, "step": 6607 }, { "epoch": 0.8406055209260909, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.003582000732422, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8600101470947266, "num_tokens": 252161636.0, "step": 6608 }, { "epoch": 0.8407327312046813, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.055218696594238, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8631483316421509, "num_tokens": 252199669.0, "step": 6609 }, { "epoch": 0.8408599414832718, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.995460510253906, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8701083660125732, "num_tokens": 252236832.0, "step": 6610 }, { "epoch": 0.8409871517618623, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.021918296813965, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.858228325843811, "num_tokens": 252282518.0, "step": 6611 }, { "epoch": 0.8411143620404529, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.0853910446167, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8562337160110474, "num_tokens": 252316479.0, "step": 6612 }, { "epoch": 0.8412415723190434, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.984583854675293, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8566175699234009, "num_tokens": 252353789.0, "step": 6613 }, { "epoch": 0.8413687825976339, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.044265747070312, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8660233020782471, "num_tokens": 252391609.0, "step": 6614 }, { "epoch": 0.8414959928762245, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.9444580078125, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8753510117530823, "num_tokens": 252433514.0, "step": 6615 }, { "epoch": 0.8416232031548149, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.993840217590332, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8665382862091064, "num_tokens": 252469700.0, "step": 6616 }, { "epoch": 0.8417504134334054, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.03336238861084, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.858802855014801, "num_tokens": 252501658.0, "step": 6617 }, { "epoch": 0.8418776237119959, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.046673774719238, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8647847175598145, "num_tokens": 252534546.0, "step": 6618 }, { "epoch": 0.8420048339905865, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00610637664795, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8705968856811523, "num_tokens": 252573172.0, "step": 6619 }, { "epoch": 0.842132044269177, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.990300178527832, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.871976375579834, "num_tokens": 252607258.0, "step": 6620 }, { "epoch": 0.8422592545477675, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.985552787780762, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8730166554450989, "num_tokens": 252651451.0, "step": 6621 }, { "epoch": 0.8423864648263579, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.99682331085205, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8738359212875366, "num_tokens": 252686348.0, "step": 6622 }, { "epoch": 0.8425136751049485, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.021053314208984, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8567312955856323, "num_tokens": 252724316.0, "step": 6623 }, { "epoch": 0.842640885383539, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.053796768188477, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8517099618911743, "num_tokens": 252764612.0, "step": 6624 }, { "epoch": 0.8427680956621295, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.053308486938477, "learning_rate": 1e-06, "loss": 0.4892, "mean_token_accuracy": 0.841205358505249, "num_tokens": 252809510.0, "step": 6625 }, { "epoch": 0.84289530594072, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.03549575805664, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8597800731658936, "num_tokens": 252847598.0, "step": 6626 }, { "epoch": 0.8430225162193106, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.02053451538086, "learning_rate": 1e-06, "loss": 0.4899, "mean_token_accuracy": 0.848386824131012, "num_tokens": 252884401.0, "step": 6627 }, { "epoch": 0.843149726497901, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.030941009521484, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8467027544975281, "num_tokens": 252928348.0, "step": 6628 }, { "epoch": 0.8432769367764915, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.063485145568848, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8622817397117615, "num_tokens": 252965084.0, "step": 6629 }, { "epoch": 0.843404147055082, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.029561042785645, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8666699528694153, "num_tokens": 253003697.0, "step": 6630 }, { "epoch": 0.8435313573336726, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.113393783569336, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8489764928817749, "num_tokens": 253037239.0, "step": 6631 }, { "epoch": 0.8436585676122631, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.031139373779297, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.853506326675415, "num_tokens": 253078163.0, "step": 6632 }, { "epoch": 0.8437857778908536, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.005515098571777, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8604692220687866, "num_tokens": 253118907.0, "step": 6633 }, { "epoch": 0.843912988169444, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.089813232421875, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8808058500289917, "num_tokens": 253153132.0, "step": 6634 }, { "epoch": 0.8440401984480346, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.065641403198242, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8590197563171387, "num_tokens": 253194446.0, "step": 6635 }, { "epoch": 0.8441674087266251, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.102498054504395, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.86489337682724, "num_tokens": 253230798.0, "step": 6636 }, { "epoch": 0.8442946190052156, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.056330680847168, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8553882241249084, "num_tokens": 253269029.0, "step": 6637 }, { "epoch": 0.8444218292838062, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.054041862487793, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8575069904327393, "num_tokens": 253310656.0, "step": 6638 }, { "epoch": 0.8445490395623967, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.0631742477417, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8801870942115784, "num_tokens": 253346882.0, "step": 6639 }, { "epoch": 0.8446762498409871, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.06309986114502, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8530603051185608, "num_tokens": 253389601.0, "step": 6640 }, { "epoch": 0.8448034601195776, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.013301849365234, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8613977432250977, "num_tokens": 253428725.0, "step": 6641 }, { "epoch": 0.8449306703981682, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.096741676330566, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8764585256576538, "num_tokens": 253461404.0, "step": 6642 }, { "epoch": 0.8450578806767587, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.01285171508789, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.866925060749054, "num_tokens": 253500324.0, "step": 6643 }, { "epoch": 0.8451850909553492, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.052101135253906, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8500341176986694, "num_tokens": 253542500.0, "step": 6644 }, { "epoch": 0.8453123012339397, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.024530410766602, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8450664281845093, "num_tokens": 253581887.0, "step": 6645 }, { "epoch": 0.8454395115125302, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.10571575164795, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8706496357917786, "num_tokens": 253616942.0, "step": 6646 }, { "epoch": 0.8455667217911207, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.023381233215332, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.856078028678894, "num_tokens": 253652610.0, "step": 6647 }, { "epoch": 0.8456939320697112, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.070488929748535, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8593068718910217, "num_tokens": 253686269.0, "step": 6648 }, { "epoch": 0.8458211423483017, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.08851146697998, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.85745769739151, "num_tokens": 253723632.0, "step": 6649 }, { "epoch": 0.8459483526268923, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.08871078491211, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8561906814575195, "num_tokens": 253755990.0, "step": 6650 }, { "epoch": 0.8460755629054828, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.096054077148438, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.857114315032959, "num_tokens": 253794200.0, "step": 6651 }, { "epoch": 0.8462027731840732, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.006194114685059, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8730648756027222, "num_tokens": 253840554.0, "step": 6652 }, { "epoch": 0.8463299834626637, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.062920570373535, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8721995949745178, "num_tokens": 253873642.0, "step": 6653 }, { "epoch": 0.8464571937412543, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.0713472366333, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8677788972854614, "num_tokens": 253911839.0, "step": 6654 }, { "epoch": 0.8465844040198448, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.038407325744629, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8526161313056946, "num_tokens": 253944649.0, "step": 6655 }, { "epoch": 0.8467116142984353, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 28.714834213256836, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8616786599159241, "num_tokens": 253977739.0, "step": 6656 }, { "epoch": 0.8468388245770259, "ewc_loss": 0.0322265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.412827491760254, "learning_rate": 1e-06, "loss": 0.4756, "mean_token_accuracy": 0.8473654389381409, "num_tokens": 254021216.0, "step": 6657 }, { "epoch": 0.8469660348556163, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 9.657808303833008, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8598607778549194, "num_tokens": 254063209.0, "step": 6658 }, { "epoch": 0.8470932451342068, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 12.765885353088379, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8580151796340942, "num_tokens": 254102325.0, "step": 6659 }, { "epoch": 0.8472204554127973, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.561361312866211, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8666077852249146, "num_tokens": 254136991.0, "step": 6660 }, { "epoch": 0.8473476656913879, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.338444709777832, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8621342182159424, "num_tokens": 254179520.0, "step": 6661 }, { "epoch": 0.8474748759699784, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.71311092376709, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8622555732727051, "num_tokens": 254220661.0, "step": 6662 }, { "epoch": 0.8476020862485689, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 11.7052583694458, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8700185418128967, "num_tokens": 254261853.0, "step": 6663 }, { "epoch": 0.8477292965271594, "ewc_loss": 0.025146484375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.517979621887207, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8630017042160034, "num_tokens": 254301577.0, "step": 6664 }, { "epoch": 0.8478565068057499, "ewc_loss": 0.02490234375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.35469913482666, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8752605319023132, "num_tokens": 254341792.0, "step": 6665 }, { "epoch": 0.8479837170843404, "ewc_loss": 0.0244140625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.387578010559082, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.849769115447998, "num_tokens": 254376755.0, "step": 6666 }, { "epoch": 0.8481109273629309, "ewc_loss": 0.0244140625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.286697387695312, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8738066554069519, "num_tokens": 254413565.0, "step": 6667 }, { "epoch": 0.8482381376415215, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.308334350585938, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8522339463233948, "num_tokens": 254449889.0, "step": 6668 }, { "epoch": 0.848365347920112, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.192817687988281, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8767329454421997, "num_tokens": 254488129.0, "step": 6669 }, { "epoch": 0.8484925581987025, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.156944274902344, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8731801509857178, "num_tokens": 254522031.0, "step": 6670 }, { "epoch": 0.8486197684772929, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.146793365478516, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8732256293296814, "num_tokens": 254559249.0, "step": 6671 }, { "epoch": 0.8487469787558835, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.113070487976074, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8556191325187683, "num_tokens": 254603968.0, "step": 6672 }, { "epoch": 0.848874189034474, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.094286918640137, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8713556528091431, "num_tokens": 254642178.0, "step": 6673 }, { "epoch": 0.8490013993130645, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.061564445495605, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.854767918586731, "num_tokens": 254679129.0, "step": 6674 }, { "epoch": 0.849128609591655, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.140167236328125, "learning_rate": 1e-06, "loss": 0.5319, "mean_token_accuracy": 0.8335245847702026, "num_tokens": 254714952.0, "step": 6675 }, { "epoch": 0.8492558198702456, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.057831764221191, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8582709431648254, "num_tokens": 254743323.0, "step": 6676 }, { "epoch": 0.849383030148836, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.037029266357422, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8686478137969971, "num_tokens": 254784749.0, "step": 6677 }, { "epoch": 0.8495102404274265, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.098857879638672, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8677294254302979, "num_tokens": 254820102.0, "step": 6678 }, { "epoch": 0.849637450706017, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.065589904785156, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.858362078666687, "num_tokens": 254859988.0, "step": 6679 }, { "epoch": 0.8497646609846076, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.972455978393555, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8580504059791565, "num_tokens": 254906792.0, "step": 6680 }, { "epoch": 0.8498918712631981, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.065115928649902, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8640979528427124, "num_tokens": 254941552.0, "step": 6681 }, { "epoch": 0.8500190815417886, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.03099536895752, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8542813658714294, "num_tokens": 254974653.0, "step": 6682 }, { "epoch": 0.850146291820379, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.00981616973877, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.847606897354126, "num_tokens": 255018418.0, "step": 6683 }, { "epoch": 0.8502735020989696, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.996106147766113, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8674215078353882, "num_tokens": 255049626.0, "step": 6684 }, { "epoch": 0.8504007123775601, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.04310417175293, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8557590246200562, "num_tokens": 255086323.0, "step": 6685 }, { "epoch": 0.8505279226561506, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.026688575744629, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8699892163276672, "num_tokens": 255124102.0, "step": 6686 }, { "epoch": 0.8506551329347412, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.988412857055664, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8660179376602173, "num_tokens": 255162176.0, "step": 6687 }, { "epoch": 0.8507823432133317, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.053504943847656, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8652142286300659, "num_tokens": 255196256.0, "step": 6688 }, { "epoch": 0.8509095534919221, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.009435653686523, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8552550077438354, "num_tokens": 255236752.0, "step": 6689 }, { "epoch": 0.8510367637705126, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.059514999389648, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.863345742225647, "num_tokens": 255268694.0, "step": 6690 }, { "epoch": 0.8511639740491032, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.051258087158203, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8682702779769897, "num_tokens": 255309445.0, "step": 6691 }, { "epoch": 0.8512911843276937, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.053481101989746, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8650290369987488, "num_tokens": 255341126.0, "step": 6692 }, { "epoch": 0.8514183946062842, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.0035982131958, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8715426921844482, "num_tokens": 255379553.0, "step": 6693 }, { "epoch": 0.8515456048848747, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.953896522521973, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8631951808929443, "num_tokens": 255416588.0, "step": 6694 }, { "epoch": 0.8516728151634652, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.068714141845703, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8505103588104248, "num_tokens": 255449574.0, "step": 6695 }, { "epoch": 0.8518000254420557, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.974624633789062, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8602538108825684, "num_tokens": 255494088.0, "step": 6696 }, { "epoch": 0.8519272357206462, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.046060562133789, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8715076446533203, "num_tokens": 255529785.0, "step": 6697 }, { "epoch": 0.8520544459992367, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.020288467407227, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.848903477191925, "num_tokens": 255566675.0, "step": 6698 }, { "epoch": 0.8521816562778273, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.013066291809082, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8672762513160706, "num_tokens": 255606502.0, "step": 6699 }, { "epoch": 0.8523088665564178, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.023041725158691, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8533077239990234, "num_tokens": 255649179.0, "step": 6700 }, { "epoch": 0.8524360768350082, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.990917205810547, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8612433671951294, "num_tokens": 255688190.0, "step": 6701 }, { "epoch": 0.8525632871135987, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.033292770385742, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8601855039596558, "num_tokens": 255730401.0, "step": 6702 }, { "epoch": 0.8526904973921893, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.013311386108398, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8581579923629761, "num_tokens": 255767881.0, "step": 6703 }, { "epoch": 0.8528177076707798, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.017704963684082, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.855492353439331, "num_tokens": 255808347.0, "step": 6704 }, { "epoch": 0.8529449179493703, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.002616882324219, "learning_rate": 1e-06, "loss": 0.4914, "mean_token_accuracy": 0.8435415029525757, "num_tokens": 255849538.0, "step": 6705 }, { "epoch": 0.8530721282279609, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.974839210510254, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8727792501449585, "num_tokens": 255890228.0, "step": 6706 }, { "epoch": 0.8531993385065513, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.995611190795898, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8604351282119751, "num_tokens": 255930189.0, "step": 6707 }, { "epoch": 0.8533265487851418, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.994941711425781, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8687704205513, "num_tokens": 255968310.0, "step": 6708 }, { "epoch": 0.8534537590637323, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.965073585510254, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.873482882976532, "num_tokens": 256005872.0, "step": 6709 }, { "epoch": 0.8535809693423229, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.020729064941406, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8674886226654053, "num_tokens": 256045374.0, "step": 6710 }, { "epoch": 0.8537081796209134, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.97783374786377, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.865649938583374, "num_tokens": 256084874.0, "step": 6711 }, { "epoch": 0.8538353898995039, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.009509086608887, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8553996086120605, "num_tokens": 256122251.0, "step": 6712 }, { "epoch": 0.8539626001780944, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.004803657531738, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8597676753997803, "num_tokens": 256165025.0, "step": 6713 }, { "epoch": 0.8540898104566849, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.977180480957031, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8686405420303345, "num_tokens": 256202220.0, "step": 6714 }, { "epoch": 0.8542170207352754, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.97726821899414, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8647986650466919, "num_tokens": 256234036.0, "step": 6715 }, { "epoch": 0.8543442310138659, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.994098663330078, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8605378866195679, "num_tokens": 256275253.0, "step": 6716 }, { "epoch": 0.8544714412924564, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.020962715148926, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8597704172134399, "num_tokens": 256311296.0, "step": 6717 }, { "epoch": 0.854598651571047, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.026298522949219, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8687241077423096, "num_tokens": 256348149.0, "step": 6718 }, { "epoch": 0.8547258618496375, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.95108413696289, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8763518333435059, "num_tokens": 256387861.0, "step": 6719 }, { "epoch": 0.8548530721282279, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.045955657958984, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.871907114982605, "num_tokens": 256427131.0, "step": 6720 }, { "epoch": 0.8549802824068184, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.99429988861084, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8671000599861145, "num_tokens": 256465041.0, "step": 6721 }, { "epoch": 0.855107492685409, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.998405456542969, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8541688323020935, "num_tokens": 256502576.0, "step": 6722 }, { "epoch": 0.8552347029639995, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.035147666931152, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8659251928329468, "num_tokens": 256540493.0, "step": 6723 }, { "epoch": 0.85536191324259, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.042205810546875, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8452363014221191, "num_tokens": 256578115.0, "step": 6724 }, { "epoch": 0.8554891235211806, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.985946655273438, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8703848719596863, "num_tokens": 256615957.0, "step": 6725 }, { "epoch": 0.855616333799771, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.02513313293457, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8655258417129517, "num_tokens": 256649972.0, "step": 6726 }, { "epoch": 0.8557435440783615, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.02792739868164, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8700664043426514, "num_tokens": 256686683.0, "step": 6727 }, { "epoch": 0.855870754356952, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.990687370300293, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8609628081321716, "num_tokens": 256723587.0, "step": 6728 }, { "epoch": 0.8559979646355426, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.049843788146973, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8629664778709412, "num_tokens": 256765245.0, "step": 6729 }, { "epoch": 0.8561251749141331, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.94633674621582, "learning_rate": 1e-06, "loss": 0.3333, "mean_token_accuracy": 0.8898045420646667, "num_tokens": 256806930.0, "step": 6730 }, { "epoch": 0.8562523851927236, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.025217056274414, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8606181144714355, "num_tokens": 256844568.0, "step": 6731 }, { "epoch": 0.856379595471314, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.004278182983398, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8610588312149048, "num_tokens": 256877483.0, "step": 6732 }, { "epoch": 0.8565068057499046, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.991756439208984, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8673078417778015, "num_tokens": 256921139.0, "step": 6733 }, { "epoch": 0.8566340160284951, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.970271110534668, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.863063633441925, "num_tokens": 256957861.0, "step": 6734 }, { "epoch": 0.8567612263070856, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.005512237548828, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8645508289337158, "num_tokens": 256998582.0, "step": 6735 }, { "epoch": 0.8568884365856761, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.992445945739746, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.862087070941925, "num_tokens": 257037638.0, "step": 6736 }, { "epoch": 0.8570156468642667, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.020050048828125, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8643268346786499, "num_tokens": 257074427.0, "step": 6737 }, { "epoch": 0.8571428571428571, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.937506675720215, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8654618263244629, "num_tokens": 257113826.0, "step": 6738 }, { "epoch": 0.8572700674214476, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.985301971435547, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8491479158401489, "num_tokens": 257152387.0, "step": 6739 }, { "epoch": 0.8573972777000382, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.982718467712402, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8638913631439209, "num_tokens": 257192083.0, "step": 6740 }, { "epoch": 0.8575244879786287, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.990218162536621, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8550548553466797, "num_tokens": 257235425.0, "step": 6741 }, { "epoch": 0.8576516982572192, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.101156234741211, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8588616251945496, "num_tokens": 257268621.0, "step": 6742 }, { "epoch": 0.8577789085358097, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.014490127563477, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.867148220539093, "num_tokens": 257306909.0, "step": 6743 }, { "epoch": 0.8579061188144002, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 11.021158218383789, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8771641254425049, "num_tokens": 257343388.0, "step": 6744 }, { "epoch": 0.8580333290929907, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.038301467895508, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8685981035232544, "num_tokens": 257377172.0, "step": 6745 }, { "epoch": 0.8581605393715812, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.012222290039062, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8557791709899902, "num_tokens": 257410312.0, "step": 6746 }, { "epoch": 0.8582877496501717, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.998307228088379, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8661602735519409, "num_tokens": 257449919.0, "step": 6747 }, { "epoch": 0.8584149599287623, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.003669738769531, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8736395835876465, "num_tokens": 257488296.0, "step": 6748 }, { "epoch": 0.8585421702073528, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.014054298400879, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8714733123779297, "num_tokens": 257524678.0, "step": 6749 }, { "epoch": 0.8586693804859432, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.039069175720215, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8682178258895874, "num_tokens": 257559060.0, "step": 6750 }, { "epoch": 0.8587965907645337, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.04526138305664, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.872467041015625, "num_tokens": 257594340.0, "step": 6751 }, { "epoch": 0.8589238010431243, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.031482696533203, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8746187686920166, "num_tokens": 257627936.0, "step": 6752 }, { "epoch": 0.8590510113217148, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.025399208068848, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.860919713973999, "num_tokens": 257660311.0, "step": 6753 }, { "epoch": 0.8591782216003053, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.035887718200684, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8552597761154175, "num_tokens": 257697496.0, "step": 6754 }, { "epoch": 0.8593054318788959, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.038259506225586, "learning_rate": 1e-06, "loss": 0.4801, "mean_token_accuracy": 0.8453284502029419, "num_tokens": 257737242.0, "step": 6755 }, { "epoch": 0.8594326421574863, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.059592247009277, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8636690378189087, "num_tokens": 257772946.0, "step": 6756 }, { "epoch": 0.8595598524360768, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.050832748413086, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.857641339302063, "num_tokens": 257807198.0, "step": 6757 }, { "epoch": 0.8596870627146673, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.955068588256836, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.869655966758728, "num_tokens": 257848618.0, "step": 6758 }, { "epoch": 0.8598142729932579, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.035181999206543, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8749868273735046, "num_tokens": 257887805.0, "step": 6759 }, { "epoch": 0.8599414832718484, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.060654640197754, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8686536550521851, "num_tokens": 257929276.0, "step": 6760 }, { "epoch": 0.8600686935504389, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.074615478515625, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8579001426696777, "num_tokens": 257958629.0, "step": 6761 }, { "epoch": 0.8601959038290294, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.038381576538086, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8596675992012024, "num_tokens": 257990303.0, "step": 6762 }, { "epoch": 0.8603231141076199, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.068133354187012, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.8461838960647583, "num_tokens": 258028704.0, "step": 6763 }, { "epoch": 0.8604503243862104, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.012723922729492, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8721891045570374, "num_tokens": 258062757.0, "step": 6764 }, { "epoch": 0.8605775346648009, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.037276268005371, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8629425764083862, "num_tokens": 258098574.0, "step": 6765 }, { "epoch": 0.8607047449433914, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.08780288696289, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8711943626403809, "num_tokens": 258130937.0, "step": 6766 }, { "epoch": 0.860831955221982, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.962697982788086, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8657158613204956, "num_tokens": 258165142.0, "step": 6767 }, { "epoch": 0.8609591655005725, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.0455904006958, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8798927068710327, "num_tokens": 258203986.0, "step": 6768 }, { "epoch": 0.8610863757791629, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.053592681884766, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8644344806671143, "num_tokens": 258238692.0, "step": 6769 }, { "epoch": 0.8612135860577534, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.025647163391113, "learning_rate": 1e-06, "loss": 0.4959, "mean_token_accuracy": 0.8445015549659729, "num_tokens": 258280942.0, "step": 6770 }, { "epoch": 0.861340796336344, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.055956840515137, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8671698570251465, "num_tokens": 258318835.0, "step": 6771 }, { "epoch": 0.8614680066149345, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.034065246582031, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8554291725158691, "num_tokens": 258359461.0, "step": 6772 }, { "epoch": 0.861595216893525, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.099750518798828, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8593102693557739, "num_tokens": 258394582.0, "step": 6773 }, { "epoch": 0.8617224271721156, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.020364761352539, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8666605353355408, "num_tokens": 258438526.0, "step": 6774 }, { "epoch": 0.861849637450706, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.047619819641113, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8502932786941528, "num_tokens": 258479332.0, "step": 6775 }, { "epoch": 0.8619768477292965, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.115670204162598, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8539004921913147, "num_tokens": 258513202.0, "step": 6776 }, { "epoch": 0.862104058007887, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.983025550842285, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8658343553543091, "num_tokens": 258549145.0, "step": 6777 }, { "epoch": 0.8622312682864776, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.085620880126953, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8551695346832275, "num_tokens": 258586421.0, "step": 6778 }, { "epoch": 0.8623584785650681, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.067044258117676, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8643925786018372, "num_tokens": 258620014.0, "step": 6779 }, { "epoch": 0.8624856888436586, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.050198554992676, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8680728077888489, "num_tokens": 258656124.0, "step": 6780 }, { "epoch": 0.862612899122249, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.010968208312988, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.852552592754364, "num_tokens": 258692537.0, "step": 6781 }, { "epoch": 0.8627401094008396, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.074515342712402, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8664465546607971, "num_tokens": 258732737.0, "step": 6782 }, { "epoch": 0.8628673196794301, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.02275276184082, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.850311279296875, "num_tokens": 258770634.0, "step": 6783 }, { "epoch": 0.8629945299580206, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.078503608703613, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8622692823410034, "num_tokens": 258809643.0, "step": 6784 }, { "epoch": 0.8631217402366111, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.061625480651855, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8690446615219116, "num_tokens": 258845336.0, "step": 6785 }, { "epoch": 0.8632489505152017, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.053598403930664, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8650959730148315, "num_tokens": 258882206.0, "step": 6786 }, { "epoch": 0.8633761607937921, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.035350799560547, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.870757520198822, "num_tokens": 258918661.0, "step": 6787 }, { "epoch": 0.8635033710723826, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.129789352416992, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8580262660980225, "num_tokens": 258957452.0, "step": 6788 }, { "epoch": 0.8636305813509731, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.019088745117188, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8543517589569092, "num_tokens": 259003507.0, "step": 6789 }, { "epoch": 0.8637577916295637, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.056451797485352, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8583534955978394, "num_tokens": 259042284.0, "step": 6790 }, { "epoch": 0.8638850019081542, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.024078369140625, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8767919540405273, "num_tokens": 259082929.0, "step": 6791 }, { "epoch": 0.8640122121867447, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.052473068237305, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.877964973449707, "num_tokens": 259125920.0, "step": 6792 }, { "epoch": 0.8641394224653351, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.037764549255371, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8646717667579651, "num_tokens": 259163140.0, "step": 6793 }, { "epoch": 0.8642666327439257, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.034276008605957, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8612216711044312, "num_tokens": 259199038.0, "step": 6794 }, { "epoch": 0.8643938430225162, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.063882827758789, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.861177921295166, "num_tokens": 259235713.0, "step": 6795 }, { "epoch": 0.8645210533011067, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.039104461669922, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.852765679359436, "num_tokens": 259278314.0, "step": 6796 }, { "epoch": 0.8646482635796973, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.074244499206543, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.871985912322998, "num_tokens": 259313039.0, "step": 6797 }, { "epoch": 0.8647754738582878, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.054691314697266, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8696569204330444, "num_tokens": 259348845.0, "step": 6798 }, { "epoch": 0.8649026841368782, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.012004852294922, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8572736978530884, "num_tokens": 259387345.0, "step": 6799 }, { "epoch": 0.8650298944154687, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.012580871582031, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8626618981361389, "num_tokens": 259427835.0, "step": 6800 }, { "epoch": 0.8651571046940593, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.090902328491211, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8666783571243286, "num_tokens": 259460453.0, "step": 6801 }, { "epoch": 0.8652843149726498, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.01025676727295, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8696954250335693, "num_tokens": 259499296.0, "step": 6802 }, { "epoch": 0.8654115252512403, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.043896675109863, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8605077266693115, "num_tokens": 259537649.0, "step": 6803 }, { "epoch": 0.8655387355298308, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.015387535095215, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8557083606719971, "num_tokens": 259583136.0, "step": 6804 }, { "epoch": 0.8656659458084213, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.978482246398926, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8670463562011719, "num_tokens": 259620422.0, "step": 6805 }, { "epoch": 0.8657931560870118, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.008354187011719, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8516315221786499, "num_tokens": 259663948.0, "step": 6806 }, { "epoch": 0.8659203663656023, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.051321029663086, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8648339509963989, "num_tokens": 259701986.0, "step": 6807 }, { "epoch": 0.8660475766441929, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.998274803161621, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8780959248542786, "num_tokens": 259740684.0, "step": 6808 }, { "epoch": 0.8661747869227834, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.022880554199219, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8468523621559143, "num_tokens": 259781659.0, "step": 6809 }, { "epoch": 0.8663019972013739, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.015523910522461, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8578911423683167, "num_tokens": 259818554.0, "step": 6810 }, { "epoch": 0.8664292074799644, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.045790672302246, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8647648096084595, "num_tokens": 259856652.0, "step": 6811 }, { "epoch": 0.8665564177585549, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.026412963867188, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8805974125862122, "num_tokens": 259898255.0, "step": 6812 }, { "epoch": 0.8666836280371454, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.03943157196045, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8683615326881409, "num_tokens": 259935305.0, "step": 6813 }, { "epoch": 0.8668108383157359, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.0634765625, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8683789968490601, "num_tokens": 259975245.0, "step": 6814 }, { "epoch": 0.8669380485943264, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.082837104797363, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8679741621017456, "num_tokens": 260014150.0, "step": 6815 }, { "epoch": 0.867065258872917, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.06078815460205, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8537530899047852, "num_tokens": 260054655.0, "step": 6816 }, { "epoch": 0.8671924691515075, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.156317710876465, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8634053468704224, "num_tokens": 260091528.0, "step": 6817 }, { "epoch": 0.8673196794300979, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.044829368591309, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8790569305419922, "num_tokens": 260123069.0, "step": 6818 }, { "epoch": 0.8674468897086884, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.192657470703125, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8483569622039795, "num_tokens": 260154482.0, "step": 6819 }, { "epoch": 0.867574099987279, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.072555541992188, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8652772903442383, "num_tokens": 260193983.0, "step": 6820 }, { "epoch": 0.8677013102658695, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.033232688903809, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8652137517929077, "num_tokens": 260228705.0, "step": 6821 }, { "epoch": 0.86782852054446, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.193662643432617, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8620527982711792, "num_tokens": 260262821.0, "step": 6822 }, { "epoch": 0.8679557308230506, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.035260200500488, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8614929914474487, "num_tokens": 260298949.0, "step": 6823 }, { "epoch": 0.868082941101641, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.039590835571289, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8707051873207092, "num_tokens": 260340028.0, "step": 6824 }, { "epoch": 0.8682101513802315, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.09307861328125, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8635094165802002, "num_tokens": 260381328.0, "step": 6825 }, { "epoch": 0.868337361658822, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.057921409606934, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8639441728591919, "num_tokens": 260423421.0, "step": 6826 }, { "epoch": 0.8684645719374126, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.087986946105957, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8624107837677002, "num_tokens": 260461985.0, "step": 6827 }, { "epoch": 0.8685917822160031, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.061156272888184, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8477060198783875, "num_tokens": 260503041.0, "step": 6828 }, { "epoch": 0.8687189924945936, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.095598220825195, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8717994689941406, "num_tokens": 260545553.0, "step": 6829 }, { "epoch": 0.868846202773184, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.102269172668457, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.871060311794281, "num_tokens": 260577160.0, "step": 6830 }, { "epoch": 0.8689734130517746, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.11264419555664, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8743664026260376, "num_tokens": 260614748.0, "step": 6831 }, { "epoch": 0.8691006233303651, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.08874225616455, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8634487986564636, "num_tokens": 260651762.0, "step": 6832 }, { "epoch": 0.8692278336089556, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.099690437316895, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8520789742469788, "num_tokens": 260694438.0, "step": 6833 }, { "epoch": 0.8693550438875461, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.079412460327148, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8721274733543396, "num_tokens": 260730256.0, "step": 6834 }, { "epoch": 0.8694822541661367, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.11883544921875, "learning_rate": 1e-06, "loss": 0.4894, "mean_token_accuracy": 0.8424792289733887, "num_tokens": 260766190.0, "step": 6835 }, { "epoch": 0.8696094644447271, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.100761413574219, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8778893947601318, "num_tokens": 260801733.0, "step": 6836 }, { "epoch": 0.8697366747233176, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.101234436035156, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8749641180038452, "num_tokens": 260843461.0, "step": 6837 }, { "epoch": 0.8698638850019081, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.020307540893555, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8613704442977905, "num_tokens": 260875006.0, "step": 6838 }, { "epoch": 0.8699910952804987, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.138450622558594, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8576371669769287, "num_tokens": 260917553.0, "step": 6839 }, { "epoch": 0.8701183055590892, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.063637733459473, "learning_rate": 1e-06, "loss": 0.4973, "mean_token_accuracy": 0.8385992050170898, "num_tokens": 260955293.0, "step": 6840 }, { "epoch": 0.8702455158376797, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.097094535827637, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.863279402256012, "num_tokens": 260983966.0, "step": 6841 }, { "epoch": 0.8703727261162701, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 11.102209091186523, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8758931159973145, "num_tokens": 261019434.0, "step": 6842 }, { "epoch": 0.8704999363948607, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.95527458190918, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8524307608604431, "num_tokens": 261053761.0, "step": 6843 }, { "epoch": 0.8706271466734512, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.090620040893555, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8537441492080688, "num_tokens": 261093490.0, "step": 6844 }, { "epoch": 0.8707543569520417, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.105154037475586, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8650777339935303, "num_tokens": 261129518.0, "step": 6845 }, { "epoch": 0.8708815672306323, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.088846206665039, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8466646671295166, "num_tokens": 261164282.0, "step": 6846 }, { "epoch": 0.8710087775092228, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.076093673706055, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.8434709906578064, "num_tokens": 261205055.0, "step": 6847 }, { "epoch": 0.8711359877878132, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.113481521606445, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8579006195068359, "num_tokens": 261242329.0, "step": 6848 }, { "epoch": 0.8712631980664037, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.103750228881836, "learning_rate": 1e-06, "loss": 0.5148, "mean_token_accuracy": 0.8368459343910217, "num_tokens": 261280330.0, "step": 6849 }, { "epoch": 0.8713904083449943, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.092581748962402, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8739566802978516, "num_tokens": 261318232.0, "step": 6850 }, { "epoch": 0.8715176186235848, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.135627746582031, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8797216415405273, "num_tokens": 261351483.0, "step": 6851 }, { "epoch": 0.8716448289021753, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.096566200256348, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8656169176101685, "num_tokens": 261387058.0, "step": 6852 }, { "epoch": 0.8717720391807658, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.0452299118042, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8780294060707092, "num_tokens": 261430110.0, "step": 6853 }, { "epoch": 0.8718992494593563, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.047126770019531, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8697593212127686, "num_tokens": 261470707.0, "step": 6854 }, { "epoch": 0.8720264597379468, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.078383445739746, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8706323504447937, "num_tokens": 261507146.0, "step": 6855 }, { "epoch": 0.8721536700165373, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.075151443481445, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8493896722793579, "num_tokens": 261549723.0, "step": 6856 }, { "epoch": 0.8722808802951278, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.066815376281738, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8688730597496033, "num_tokens": 261591030.0, "step": 6857 }, { "epoch": 0.8724080905737184, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.05504035949707, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8729515075683594, "num_tokens": 261623919.0, "step": 6858 }, { "epoch": 0.8725353008523089, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.093904495239258, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8806328773498535, "num_tokens": 261655277.0, "step": 6859 }, { "epoch": 0.8726625111308994, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.04598331451416, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.8507838845252991, "num_tokens": 261697462.0, "step": 6860 }, { "epoch": 0.8727897214094898, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.054742813110352, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8614953756332397, "num_tokens": 261739134.0, "step": 6861 }, { "epoch": 0.8729169316880804, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.02906608581543, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8667467832565308, "num_tokens": 261776295.0, "step": 6862 }, { "epoch": 0.8730441419666709, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.060962677001953, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8677289485931396, "num_tokens": 261812388.0, "step": 6863 }, { "epoch": 0.8731713522452614, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.031163215637207, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8642196655273438, "num_tokens": 261855537.0, "step": 6864 }, { "epoch": 0.873298562523852, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.030900955200195, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8722973465919495, "num_tokens": 261899922.0, "step": 6865 }, { "epoch": 0.8734257728024425, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.115063667297363, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8755253553390503, "num_tokens": 261934475.0, "step": 6866 }, { "epoch": 0.8735529830810329, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.024696350097656, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8643229007720947, "num_tokens": 261973395.0, "step": 6867 }, { "epoch": 0.8736801933596234, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.117439270019531, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.847437858581543, "num_tokens": 262011644.0, "step": 6868 }, { "epoch": 0.873807403638214, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.060721397399902, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8588206768035889, "num_tokens": 262044038.0, "step": 6869 }, { "epoch": 0.8739346139168045, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.086588859558105, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8624886870384216, "num_tokens": 262081023.0, "step": 6870 }, { "epoch": 0.874061824195395, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.087130546569824, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8575857877731323, "num_tokens": 262123981.0, "step": 6871 }, { "epoch": 0.8741890344739855, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.055610656738281, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8533061146736145, "num_tokens": 262161542.0, "step": 6872 }, { "epoch": 0.874316244752576, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.113532066345215, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.85292649269104, "num_tokens": 262202482.0, "step": 6873 }, { "epoch": 0.8744434550311665, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.089969635009766, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8654908537864685, "num_tokens": 262239499.0, "step": 6874 }, { "epoch": 0.874570665309757, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.104543685913086, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8644617795944214, "num_tokens": 262281324.0, "step": 6875 }, { "epoch": 0.8746978755883476, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.022669792175293, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8552442789077759, "num_tokens": 262318772.0, "step": 6876 }, { "epoch": 0.8748250858669381, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.123217582702637, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8773432374000549, "num_tokens": 262353264.0, "step": 6877 }, { "epoch": 0.8749522961455286, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.023619651794434, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8611570596694946, "num_tokens": 262393360.0, "step": 6878 }, { "epoch": 0.875079506424119, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.129898071289062, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8743695020675659, "num_tokens": 262432010.0, "step": 6879 }, { "epoch": 0.8752067167027096, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.031620025634766, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8630109429359436, "num_tokens": 262468828.0, "step": 6880 }, { "epoch": 0.8753339269813001, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.03513240814209, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8704012632369995, "num_tokens": 262515141.0, "step": 6881 }, { "epoch": 0.8754611372598906, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.118517875671387, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8610265254974365, "num_tokens": 262553137.0, "step": 6882 }, { "epoch": 0.8755883475384811, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.032093048095703, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8862344026565552, "num_tokens": 262593847.0, "step": 6883 }, { "epoch": 0.8757155578170717, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.075757026672363, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8534581065177917, "num_tokens": 262635048.0, "step": 6884 }, { "epoch": 0.8758427680956621, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.115618705749512, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8567029237747192, "num_tokens": 262671200.0, "step": 6885 }, { "epoch": 0.8759699783742526, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.068045616149902, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8654443025588989, "num_tokens": 262708304.0, "step": 6886 }, { "epoch": 0.8760971886528431, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.113895416259766, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8606587648391724, "num_tokens": 262750680.0, "step": 6887 }, { "epoch": 0.8762243989314337, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.054449081420898, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8588331341743469, "num_tokens": 262784845.0, "step": 6888 }, { "epoch": 0.8763516092100242, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.086609840393066, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8632309436798096, "num_tokens": 262822646.0, "step": 6889 }, { "epoch": 0.8764788194886147, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.099129676818848, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8595338463783264, "num_tokens": 262863663.0, "step": 6890 }, { "epoch": 0.8766060297672051, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.03815746307373, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8500162363052368, "num_tokens": 262903726.0, "step": 6891 }, { "epoch": 0.8767332400457957, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.100510597229004, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8524797558784485, "num_tokens": 262942748.0, "step": 6892 }, { "epoch": 0.8768604503243862, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.099640846252441, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8433828353881836, "num_tokens": 262977639.0, "step": 6893 }, { "epoch": 0.8769876606029767, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.051874160766602, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8665167093276978, "num_tokens": 263015970.0, "step": 6894 }, { "epoch": 0.8771148708815673, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.104674339294434, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8729406595230103, "num_tokens": 263058618.0, "step": 6895 }, { "epoch": 0.8772420811601578, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.042682647705078, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.84685879945755, "num_tokens": 263098628.0, "step": 6896 }, { "epoch": 0.8773692914387482, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.097443580627441, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.866915225982666, "num_tokens": 263136793.0, "step": 6897 }, { "epoch": 0.8774965017173387, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.074807167053223, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8655896782875061, "num_tokens": 263170996.0, "step": 6898 }, { "epoch": 0.8776237119959293, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.147416114807129, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8542224168777466, "num_tokens": 263199961.0, "step": 6899 }, { "epoch": 0.8777509222745198, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.04843807220459, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8621821403503418, "num_tokens": 263241541.0, "step": 6900 }, { "epoch": 0.8778781325531103, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.040663719177246, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8752790689468384, "num_tokens": 263278778.0, "step": 6901 }, { "epoch": 0.8780053428317008, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.179675102233887, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8660709857940674, "num_tokens": 263310155.0, "step": 6902 }, { "epoch": 0.8781325531102913, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00790023803711, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8597792387008667, "num_tokens": 263355793.0, "step": 6903 }, { "epoch": 0.8782597633888818, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.09129524230957, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8735008835792542, "num_tokens": 263394446.0, "step": 6904 }, { "epoch": 0.8783869736674723, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.078877449035645, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8670331239700317, "num_tokens": 263438479.0, "step": 6905 }, { "epoch": 0.8785141839460628, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.127108573913574, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8498904705047607, "num_tokens": 263478514.0, "step": 6906 }, { "epoch": 0.8786413942246534, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.089280128479004, "learning_rate": 1e-06, "loss": 0.498, "mean_token_accuracy": 0.8437414169311523, "num_tokens": 263516734.0, "step": 6907 }, { "epoch": 0.8787686045032439, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.096260070800781, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8534069061279297, "num_tokens": 263560105.0, "step": 6908 }, { "epoch": 0.8788958147818343, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.051822662353516, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8657304048538208, "num_tokens": 263597033.0, "step": 6909 }, { "epoch": 0.8790230250604248, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.088563919067383, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8701372742652893, "num_tokens": 263631106.0, "step": 6910 }, { "epoch": 0.8791502353390154, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.05794906616211, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8570613265037537, "num_tokens": 263668227.0, "step": 6911 }, { "epoch": 0.8792774456176059, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.069747924804688, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8654999732971191, "num_tokens": 263704335.0, "step": 6912 }, { "epoch": 0.8794046558961964, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.148153305053711, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8703034520149231, "num_tokens": 263742532.0, "step": 6913 }, { "epoch": 0.879531866174787, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.050750732421875, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8642394542694092, "num_tokens": 263782321.0, "step": 6914 }, { "epoch": 0.8796590764533775, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.10115909576416, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8718833923339844, "num_tokens": 263818037.0, "step": 6915 }, { "epoch": 0.8797862867319679, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.123537063598633, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8487756252288818, "num_tokens": 263854901.0, "step": 6916 }, { "epoch": 0.8799134970105584, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.103198051452637, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8680667281150818, "num_tokens": 263896241.0, "step": 6917 }, { "epoch": 0.880040707289149, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.107909202575684, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8649453520774841, "num_tokens": 263934673.0, "step": 6918 }, { "epoch": 0.8801679175677395, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.12702751159668, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8547011613845825, "num_tokens": 263973456.0, "step": 6919 }, { "epoch": 0.88029512784633, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.1455659866333, "learning_rate": 1e-06, "loss": 0.4862, "mean_token_accuracy": 0.8455411195755005, "num_tokens": 264006050.0, "step": 6920 }, { "epoch": 0.8804223381249205, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.100534439086914, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8733142614364624, "num_tokens": 264042736.0, "step": 6921 }, { "epoch": 0.880549548403511, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.076604843139648, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8543342351913452, "num_tokens": 264084433.0, "step": 6922 }, { "epoch": 0.8806767586821015, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.130219459533691, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8706731200218201, "num_tokens": 264118331.0, "step": 6923 }, { "epoch": 0.880803968960692, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.101888656616211, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8658260107040405, "num_tokens": 264152015.0, "step": 6924 }, { "epoch": 0.8809311792392825, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.068578720092773, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8698573112487793, "num_tokens": 264187085.0, "step": 6925 }, { "epoch": 0.8810583895178731, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.163189888000488, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8589243292808533, "num_tokens": 264225558.0, "step": 6926 }, { "epoch": 0.8811855997964636, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.11231517791748, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8540244698524475, "num_tokens": 264264220.0, "step": 6927 }, { "epoch": 0.881312810075054, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.137245178222656, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8665099143981934, "num_tokens": 264302097.0, "step": 6928 }, { "epoch": 0.8814400203536445, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.126980781555176, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8661885261535645, "num_tokens": 264341966.0, "step": 6929 }, { "epoch": 0.8815672306322351, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.092574119567871, "learning_rate": 1e-06, "loss": 0.5081, "mean_token_accuracy": 0.8413394689559937, "num_tokens": 264384298.0, "step": 6930 }, { "epoch": 0.8816944409108256, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.096242904663086, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8635828495025635, "num_tokens": 264422513.0, "step": 6931 }, { "epoch": 0.8818216511894161, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.040472030639648, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8759603500366211, "num_tokens": 264461819.0, "step": 6932 }, { "epoch": 0.8819488614680067, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.123032569885254, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8532885313034058, "num_tokens": 264502618.0, "step": 6933 }, { "epoch": 0.8820760717465971, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.060250282287598, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8564026951789856, "num_tokens": 264534382.0, "step": 6934 }, { "epoch": 0.8822032820251876, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.079890251159668, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8535945415496826, "num_tokens": 264575297.0, "step": 6935 }, { "epoch": 0.8823304923037781, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.047510147094727, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8698282241821289, "num_tokens": 264611822.0, "step": 6936 }, { "epoch": 0.8824577025823687, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.023882865905762, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8595462441444397, "num_tokens": 264657483.0, "step": 6937 }, { "epoch": 0.8825849128609592, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.147544860839844, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8692440986633301, "num_tokens": 264690633.0, "step": 6938 }, { "epoch": 0.8827121231395497, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.035130500793457, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8779169321060181, "num_tokens": 264733268.0, "step": 6939 }, { "epoch": 0.8828393334181401, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.191040992736816, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8568567037582397, "num_tokens": 264774769.0, "step": 6940 }, { "epoch": 0.8829665436967307, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.141035079956055, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8593282103538513, "num_tokens": 264806658.0, "step": 6941 }, { "epoch": 0.8830937539753212, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.10534954071045, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8516436815261841, "num_tokens": 264841054.0, "step": 6942 }, { "epoch": 0.8832209642539117, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.132526397705078, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8581539392471313, "num_tokens": 264880241.0, "step": 6943 }, { "epoch": 0.8833481745325023, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.046529769897461, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.865435779094696, "num_tokens": 264922449.0, "step": 6944 }, { "epoch": 0.8834753848110928, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.03782844543457, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8634897470474243, "num_tokens": 264961605.0, "step": 6945 }, { "epoch": 0.8836025950896832, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.145770072937012, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8711756467819214, "num_tokens": 264992924.0, "step": 6946 }, { "epoch": 0.8837298053682737, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.078076362609863, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.862947404384613, "num_tokens": 265031849.0, "step": 6947 }, { "epoch": 0.8838570156468643, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.13253402709961, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8752954006195068, "num_tokens": 265069714.0, "step": 6948 }, { "epoch": 0.8839842259254548, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.065386772155762, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8582491874694824, "num_tokens": 265103325.0, "step": 6949 }, { "epoch": 0.8841114362040453, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.164775848388672, "learning_rate": 1e-06, "loss": 0.4847, "mean_token_accuracy": 0.8418499827384949, "num_tokens": 265139133.0, "step": 6950 }, { "epoch": 0.8842386464826358, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.068523406982422, "learning_rate": 1e-06, "loss": 0.502, "mean_token_accuracy": 0.839410126209259, "num_tokens": 265183536.0, "step": 6951 }, { "epoch": 0.8843658567612263, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.043699264526367, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8549850583076477, "num_tokens": 265226288.0, "step": 6952 }, { "epoch": 0.8844930670398168, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.133313179016113, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8629311323165894, "num_tokens": 265265846.0, "step": 6953 }, { "epoch": 0.8846202773184073, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.117820739746094, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8774924874305725, "num_tokens": 265300612.0, "step": 6954 }, { "epoch": 0.8847474875969978, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.097861289978027, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8719122409820557, "num_tokens": 265341440.0, "step": 6955 }, { "epoch": 0.8848746978755884, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.125299453735352, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.853685736656189, "num_tokens": 265384854.0, "step": 6956 }, { "epoch": 0.8850019081541789, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.116226196289062, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8564389944076538, "num_tokens": 265420611.0, "step": 6957 }, { "epoch": 0.8851291184327693, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.145915031433105, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.86175537109375, "num_tokens": 265457720.0, "step": 6958 }, { "epoch": 0.8852563287113598, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.054478645324707, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8713719844818115, "num_tokens": 265493434.0, "step": 6959 }, { "epoch": 0.8853835389899504, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.130915641784668, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8646217584609985, "num_tokens": 265529015.0, "step": 6960 }, { "epoch": 0.8855107492685409, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.12098217010498, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8479459881782532, "num_tokens": 265568698.0, "step": 6961 }, { "epoch": 0.8856379595471314, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.124113082885742, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8636070489883423, "num_tokens": 265605631.0, "step": 6962 }, { "epoch": 0.885765169825722, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.12799072265625, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8644585609436035, "num_tokens": 265647071.0, "step": 6963 }, { "epoch": 0.8858923801043125, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.083344459533691, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8629760146141052, "num_tokens": 265688680.0, "step": 6964 }, { "epoch": 0.8860195903829029, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.089936256408691, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8689472675323486, "num_tokens": 265727707.0, "step": 6965 }, { "epoch": 0.8861468006614934, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.103681564331055, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8611932396888733, "num_tokens": 265765860.0, "step": 6966 }, { "epoch": 0.886274010940084, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.115097045898438, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8662031888961792, "num_tokens": 265802988.0, "step": 6967 }, { "epoch": 0.8864012212186745, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.137979507446289, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8584468364715576, "num_tokens": 265840614.0, "step": 6968 }, { "epoch": 0.886528431497265, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.084259986877441, "learning_rate": 1e-06, "loss": 0.5036, "mean_token_accuracy": 0.8368134498596191, "num_tokens": 265883032.0, "step": 6969 }, { "epoch": 0.8866556417758555, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.175857543945312, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8743194937705994, "num_tokens": 265918309.0, "step": 6970 }, { "epoch": 0.886782852054446, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.09932804107666, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8657580614089966, "num_tokens": 265956406.0, "step": 6971 }, { "epoch": 0.8869100623330365, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.186544418334961, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8721222281455994, "num_tokens": 265988413.0, "step": 6972 }, { "epoch": 0.887037272611627, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.116594314575195, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8620015382766724, "num_tokens": 266022036.0, "step": 6973 }, { "epoch": 0.8871644828902175, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.141942977905273, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.862343430519104, "num_tokens": 266053735.0, "step": 6974 }, { "epoch": 0.8872916931688081, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.11505126953125, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8779202103614807, "num_tokens": 266094198.0, "step": 6975 }, { "epoch": 0.8874189034473986, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.149226188659668, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8711429238319397, "num_tokens": 266132213.0, "step": 6976 }, { "epoch": 0.887546113725989, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.148280143737793, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8631064891815186, "num_tokens": 266168001.0, "step": 6977 }, { "epoch": 0.8876733240045795, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.21701717376709, "learning_rate": 1e-06, "loss": 0.5045, "mean_token_accuracy": 0.8349627256393433, "num_tokens": 266208771.0, "step": 6978 }, { "epoch": 0.8878005342831701, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.075522422790527, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8692628741264343, "num_tokens": 266248090.0, "step": 6979 }, { "epoch": 0.8879277445617606, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.163764953613281, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8519887924194336, "num_tokens": 266283748.0, "step": 6980 }, { "epoch": 0.8880549548403511, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.194912910461426, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8702854514122009, "num_tokens": 266316613.0, "step": 6981 }, { "epoch": 0.8881821651189417, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.075857162475586, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8651813268661499, "num_tokens": 266357515.0, "step": 6982 }, { "epoch": 0.8883093753975321, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.167863845825195, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.875548243522644, "num_tokens": 266393593.0, "step": 6983 }, { "epoch": 0.8884365856761226, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.036794662475586, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8744964599609375, "num_tokens": 266436104.0, "step": 6984 }, { "epoch": 0.8885637959547131, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.125972747802734, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8756585121154785, "num_tokens": 266472809.0, "step": 6985 }, { "epoch": 0.8886910062333037, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.141254425048828, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8732692003250122, "num_tokens": 266512726.0, "step": 6986 }, { "epoch": 0.8888182165118942, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.091134071350098, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8517884612083435, "num_tokens": 266546098.0, "step": 6987 }, { "epoch": 0.8889454267904847, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.119542121887207, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8539084196090698, "num_tokens": 266588478.0, "step": 6988 }, { "epoch": 0.8890726370690751, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.131730079650879, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8528140187263489, "num_tokens": 266628975.0, "step": 6989 }, { "epoch": 0.8891998473476657, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.09073257446289, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8721140027046204, "num_tokens": 266667533.0, "step": 6990 }, { "epoch": 0.8893270576262562, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.146086692810059, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8445533514022827, "num_tokens": 266706990.0, "step": 6991 }, { "epoch": 0.8894542679048467, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.117926597595215, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8513643741607666, "num_tokens": 266747314.0, "step": 6992 }, { "epoch": 0.8895814781834372, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.07572078704834, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8541480302810669, "num_tokens": 266791031.0, "step": 6993 }, { "epoch": 0.8897086884620278, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.14344596862793, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8622250556945801, "num_tokens": 266823236.0, "step": 6994 }, { "epoch": 0.8898358987406182, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.114336967468262, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8620380163192749, "num_tokens": 266867993.0, "step": 6995 }, { "epoch": 0.8899631090192087, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.09211254119873, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8450504541397095, "num_tokens": 266908573.0, "step": 6996 }, { "epoch": 0.8900903192977992, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.156390190124512, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.860421895980835, "num_tokens": 266951167.0, "step": 6997 }, { "epoch": 0.8902175295763898, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.112853050231934, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8617291450500488, "num_tokens": 266988976.0, "step": 6998 }, { "epoch": 0.8903447398549803, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.15457534790039, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8613420724868774, "num_tokens": 267022450.0, "step": 6999 }, { "epoch": 0.8904719501335708, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.175873756408691, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8563597798347473, "num_tokens": 267059933.0, "step": 7000 }, { "epoch": 0.8905991604121613, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.192000389099121, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8541321158409119, "num_tokens": 267092518.0, "step": 7001 }, { "epoch": 0.8907263706907518, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.027985572814941, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8582103848457336, "num_tokens": 267130008.0, "step": 7002 }, { "epoch": 0.8908535809693423, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.145447731018066, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8659422397613525, "num_tokens": 267171039.0, "step": 7003 }, { "epoch": 0.8909807912479328, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.093503952026367, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8652583956718445, "num_tokens": 267211197.0, "step": 7004 }, { "epoch": 0.8911080015265234, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.18528938293457, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8689628839492798, "num_tokens": 267247471.0, "step": 7005 }, { "epoch": 0.8912352118051139, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.220766067504883, "learning_rate": 1e-06, "loss": 0.5169, "mean_token_accuracy": 0.8358033299446106, "num_tokens": 267287276.0, "step": 7006 }, { "epoch": 0.8913624220837043, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.16958236694336, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.859503448009491, "num_tokens": 267323827.0, "step": 7007 }, { "epoch": 0.8914896323622948, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.181150436401367, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8704912662506104, "num_tokens": 267364053.0, "step": 7008 }, { "epoch": 0.8916168426408854, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.143532752990723, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8461102843284607, "num_tokens": 267404393.0, "step": 7009 }, { "epoch": 0.8917440529194759, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.210066795349121, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.863832950592041, "num_tokens": 267444589.0, "step": 7010 }, { "epoch": 0.8918712631980664, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.122854232788086, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8637655973434448, "num_tokens": 267484460.0, "step": 7011 }, { "epoch": 0.891998473476657, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.10887336730957, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8553221225738525, "num_tokens": 267517390.0, "step": 7012 }, { "epoch": 0.8921256837552475, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.181678771972656, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.863113284111023, "num_tokens": 267555813.0, "step": 7013 }, { "epoch": 0.8922528940338379, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.100471496582031, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8596316576004028, "num_tokens": 267595818.0, "step": 7014 }, { "epoch": 0.8923801043124284, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.186299324035645, "learning_rate": 1e-06, "loss": 0.5074, "mean_token_accuracy": 0.8370027542114258, "num_tokens": 267633106.0, "step": 7015 }, { "epoch": 0.892507314591019, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.101995468139648, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8649884462356567, "num_tokens": 267670166.0, "step": 7016 }, { "epoch": 0.8926345248696095, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.140067100524902, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8592602014541626, "num_tokens": 267707284.0, "step": 7017 }, { "epoch": 0.8927617351482, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.119132995605469, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8622321486473083, "num_tokens": 267742349.0, "step": 7018 }, { "epoch": 0.8928889454267905, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.145397186279297, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8670174479484558, "num_tokens": 267776003.0, "step": 7019 }, { "epoch": 0.893016155705381, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.094386100769043, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8580700159072876, "num_tokens": 267815887.0, "step": 7020 }, { "epoch": 0.8931433659839715, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.169269561767578, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.872368335723877, "num_tokens": 267855874.0, "step": 7021 }, { "epoch": 0.893270576262562, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.100085258483887, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8717644214630127, "num_tokens": 267897901.0, "step": 7022 }, { "epoch": 0.8933977865411525, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.126039505004883, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8734467029571533, "num_tokens": 267938967.0, "step": 7023 }, { "epoch": 0.8935249968197431, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.163954734802246, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8587914705276489, "num_tokens": 267975196.0, "step": 7024 }, { "epoch": 0.8936522070983336, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.084251403808594, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.875474214553833, "num_tokens": 268014407.0, "step": 7025 }, { "epoch": 0.893779417376924, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.157068252563477, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.878149151802063, "num_tokens": 268043892.0, "step": 7026 }, { "epoch": 0.8939066276555145, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.140089988708496, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8619458675384521, "num_tokens": 268079651.0, "step": 7027 }, { "epoch": 0.8940338379341051, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.138484001159668, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8657962083816528, "num_tokens": 268114021.0, "step": 7028 }, { "epoch": 0.8941610482126956, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.105371475219727, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8613138794898987, "num_tokens": 268155315.0, "step": 7029 }, { "epoch": 0.8942882584912861, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.116044998168945, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.860201358795166, "num_tokens": 268199312.0, "step": 7030 }, { "epoch": 0.8944154687698767, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.08635139465332, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8691083192825317, "num_tokens": 268236955.0, "step": 7031 }, { "epoch": 0.8945426790484671, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.072586059570312, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8595795631408691, "num_tokens": 268272183.0, "step": 7032 }, { "epoch": 0.8946698893270576, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.250288963317871, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8531110286712646, "num_tokens": 268305407.0, "step": 7033 }, { "epoch": 0.8947970996056481, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.070637702941895, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8742026090621948, "num_tokens": 268340443.0, "step": 7034 }, { "epoch": 0.8949243098842387, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.087353706359863, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8813408017158508, "num_tokens": 268375597.0, "step": 7035 }, { "epoch": 0.8950515201628292, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.13684368133545, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8564250469207764, "num_tokens": 268412872.0, "step": 7036 }, { "epoch": 0.8951787304414197, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.081698417663574, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8637528419494629, "num_tokens": 268452021.0, "step": 7037 }, { "epoch": 0.8953059407200101, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.123151779174805, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8719590306282043, "num_tokens": 268486354.0, "step": 7038 }, { "epoch": 0.8954331509986007, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.090012550354004, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8567724227905273, "num_tokens": 268519918.0, "step": 7039 }, { "epoch": 0.8955603612771912, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.143733024597168, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8551702499389648, "num_tokens": 268554259.0, "step": 7040 }, { "epoch": 0.8956875715557817, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.10078239440918, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8564877510070801, "num_tokens": 268595785.0, "step": 7041 }, { "epoch": 0.8958147818343722, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.068656921386719, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8474235534667969, "num_tokens": 268640805.0, "step": 7042 }, { "epoch": 0.8959419921129628, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.139251708984375, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8682130575180054, "num_tokens": 268682446.0, "step": 7043 }, { "epoch": 0.8960692023915532, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.092877388000488, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8715691566467285, "num_tokens": 268720262.0, "step": 7044 }, { "epoch": 0.8961964126701437, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.14931869506836, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8498599529266357, "num_tokens": 268760519.0, "step": 7045 }, { "epoch": 0.8963236229487342, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.133007049560547, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8560590744018555, "num_tokens": 268796621.0, "step": 7046 }, { "epoch": 0.8964508332273248, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.210569381713867, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8708560466766357, "num_tokens": 268830527.0, "step": 7047 }, { "epoch": 0.8965780435059153, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.165800094604492, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8670926094055176, "num_tokens": 268862112.0, "step": 7048 }, { "epoch": 0.8967052537845058, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.12399673461914, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.86100834608078, "num_tokens": 268901089.0, "step": 7049 }, { "epoch": 0.8968324640630962, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.060068130493164, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8647035956382751, "num_tokens": 268942766.0, "step": 7050 }, { "epoch": 0.8969596743416868, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.201292991638184, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8692271709442139, "num_tokens": 268978642.0, "step": 7051 }, { "epoch": 0.8970868846202773, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.132732391357422, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8687763214111328, "num_tokens": 269016048.0, "step": 7052 }, { "epoch": 0.8972140948988678, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.172096252441406, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.86380934715271, "num_tokens": 269054381.0, "step": 7053 }, { "epoch": 0.8973413051774584, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.158080101013184, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8666757345199585, "num_tokens": 269090672.0, "step": 7054 }, { "epoch": 0.8974685154560489, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.170557975769043, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8659918308258057, "num_tokens": 269130231.0, "step": 7055 }, { "epoch": 0.8975957257346393, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.181477546691895, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8672678470611572, "num_tokens": 269166282.0, "step": 7056 }, { "epoch": 0.8977229360132298, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.159067153930664, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.867133378982544, "num_tokens": 269205326.0, "step": 7057 }, { "epoch": 0.8978501462918204, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.194952011108398, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8641144037246704, "num_tokens": 269243903.0, "step": 7058 }, { "epoch": 0.8979773565704109, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.102727890014648, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8572289347648621, "num_tokens": 269288330.0, "step": 7059 }, { "epoch": 0.8981045668490014, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.179439544677734, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8534356951713562, "num_tokens": 269328731.0, "step": 7060 }, { "epoch": 0.898231777127592, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.077837944030762, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.863714337348938, "num_tokens": 269369576.0, "step": 7061 }, { "epoch": 0.8983589874061825, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.185547828674316, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8615885376930237, "num_tokens": 269407734.0, "step": 7062 }, { "epoch": 0.8984861976847729, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.165587425231934, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8429715633392334, "num_tokens": 269444042.0, "step": 7063 }, { "epoch": 0.8986134079633634, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.18996524810791, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8857532739639282, "num_tokens": 269480979.0, "step": 7064 }, { "epoch": 0.898740618241954, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.135038375854492, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8647154569625854, "num_tokens": 269515594.0, "step": 7065 }, { "epoch": 0.8988678285205445, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.178849220275879, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.865449845790863, "num_tokens": 269556974.0, "step": 7066 }, { "epoch": 0.898995038799135, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.221192359924316, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8570695519447327, "num_tokens": 269591088.0, "step": 7067 }, { "epoch": 0.8991222490777255, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.129565238952637, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8727269172668457, "num_tokens": 269627921.0, "step": 7068 }, { "epoch": 0.899249459356316, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.21652889251709, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8486276865005493, "num_tokens": 269666586.0, "step": 7069 }, { "epoch": 0.8993766696349065, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.221455574035645, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8653255701065063, "num_tokens": 269701205.0, "step": 7070 }, { "epoch": 0.899503879913497, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.103687286376953, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8764222860336304, "num_tokens": 269739982.0, "step": 7071 }, { "epoch": 0.8996310901920875, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.18992805480957, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8644812107086182, "num_tokens": 269773048.0, "step": 7072 }, { "epoch": 0.8997583004706781, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.195465087890625, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8623462915420532, "num_tokens": 269811343.0, "step": 7073 }, { "epoch": 0.8998855107492686, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.242566108703613, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8679063320159912, "num_tokens": 269851601.0, "step": 7074 }, { "epoch": 0.900012721027859, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.220392227172852, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8718762993812561, "num_tokens": 269891396.0, "step": 7075 }, { "epoch": 0.9001399313064495, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.130647659301758, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8634358644485474, "num_tokens": 269930292.0, "step": 7076 }, { "epoch": 0.9002671415850401, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.250092506408691, "learning_rate": 1e-06, "loss": 0.4944, "mean_token_accuracy": 0.836482048034668, "num_tokens": 269968464.0, "step": 7077 }, { "epoch": 0.9003943518636306, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.138322830200195, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8507086634635925, "num_tokens": 270011569.0, "step": 7078 }, { "epoch": 0.9005215621422211, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.196003913879395, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8591009378433228, "num_tokens": 270050608.0, "step": 7079 }, { "epoch": 0.9006487724208116, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.206986427307129, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8495872020721436, "num_tokens": 270096100.0, "step": 7080 }, { "epoch": 0.9007759826994021, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.11809253692627, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8734081983566284, "num_tokens": 270140915.0, "step": 7081 }, { "epoch": 0.9009031929779926, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.17436695098877, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8626945614814758, "num_tokens": 270184233.0, "step": 7082 }, { "epoch": 0.9010304032565831, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.138861656188965, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8620449304580688, "num_tokens": 270223889.0, "step": 7083 }, { "epoch": 0.9011576135351737, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.134178161621094, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.868316650390625, "num_tokens": 270266357.0, "step": 7084 }, { "epoch": 0.9012848238137642, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.270414352416992, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8710759878158569, "num_tokens": 270298252.0, "step": 7085 }, { "epoch": 0.9014120340923547, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.149353981018066, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8567864894866943, "num_tokens": 270343187.0, "step": 7086 }, { "epoch": 0.9015392443709451, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.16740608215332, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8527117967605591, "num_tokens": 270385094.0, "step": 7087 }, { "epoch": 0.9016664546495357, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.22043514251709, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8510169982910156, "num_tokens": 270421772.0, "step": 7088 }, { "epoch": 0.9017936649281262, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.088603973388672, "learning_rate": 1e-06, "loss": 0.4877, "mean_token_accuracy": 0.8450071811676025, "num_tokens": 270465033.0, "step": 7089 }, { "epoch": 0.9019208752067167, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.194890975952148, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8852927684783936, "num_tokens": 270501370.0, "step": 7090 }, { "epoch": 0.9020480854853072, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.150079727172852, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8674989342689514, "num_tokens": 270543571.0, "step": 7091 }, { "epoch": 0.9021752957638978, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.146477699279785, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8729851841926575, "num_tokens": 270580635.0, "step": 7092 }, { "epoch": 0.9023025060424882, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.195317268371582, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8764618635177612, "num_tokens": 270616534.0, "step": 7093 }, { "epoch": 0.9024297163210787, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.10663890838623, "learning_rate": 1e-06, "loss": 0.4915, "mean_token_accuracy": 0.8432571887969971, "num_tokens": 270653998.0, "step": 7094 }, { "epoch": 0.9025569265996692, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.21269702911377, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8633816242218018, "num_tokens": 270688023.0, "step": 7095 }, { "epoch": 0.9026841368782598, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.177000045776367, "learning_rate": 1e-06, "loss": 0.4868, "mean_token_accuracy": 0.8459381461143494, "num_tokens": 270721298.0, "step": 7096 }, { "epoch": 0.9028113471568503, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.188128471374512, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8677401542663574, "num_tokens": 270762182.0, "step": 7097 }, { "epoch": 0.9029385574354408, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.161468505859375, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8778039813041687, "num_tokens": 270799577.0, "step": 7098 }, { "epoch": 0.9030657677140312, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.140701293945312, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8643792867660522, "num_tokens": 270840218.0, "step": 7099 }, { "epoch": 0.9031929779926218, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.192331314086914, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8841080665588379, "num_tokens": 270872982.0, "step": 7100 }, { "epoch": 0.9033201882712123, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.215753555297852, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8675901293754578, "num_tokens": 270908381.0, "step": 7101 }, { "epoch": 0.9034473985498028, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.150460243225098, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8617329597473145, "num_tokens": 270949217.0, "step": 7102 }, { "epoch": 0.9035746088283934, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.267770767211914, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8467977046966553, "num_tokens": 270989667.0, "step": 7103 }, { "epoch": 0.9037018191069839, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.111359596252441, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8533821702003479, "num_tokens": 271030395.0, "step": 7104 }, { "epoch": 0.9038290293855743, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.265524864196777, "learning_rate": 1e-06, "loss": 0.4906, "mean_token_accuracy": 0.8458120822906494, "num_tokens": 271068239.0, "step": 7105 }, { "epoch": 0.9039562396641648, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.146252632141113, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8563144207000732, "num_tokens": 271110294.0, "step": 7106 }, { "epoch": 0.9040834499427554, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.176055908203125, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8801854848861694, "num_tokens": 271143288.0, "step": 7107 }, { "epoch": 0.9042106602213459, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.177427291870117, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8705807328224182, "num_tokens": 271181183.0, "step": 7108 }, { "epoch": 0.9043378704999364, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.20747184753418, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8623307943344116, "num_tokens": 271216201.0, "step": 7109 }, { "epoch": 0.9044650807785269, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.195086479187012, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8702681064605713, "num_tokens": 271250158.0, "step": 7110 }, { "epoch": 0.9045922910571175, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.166653633117676, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8762589693069458, "num_tokens": 271289994.0, "step": 7111 }, { "epoch": 0.9047195013357079, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.187822341918945, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8695008754730225, "num_tokens": 271325782.0, "step": 7112 }, { "epoch": 0.9048467116142984, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.2090482711792, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8503121137619019, "num_tokens": 271361604.0, "step": 7113 }, { "epoch": 0.9049739218928889, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.138073921203613, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8733661770820618, "num_tokens": 271398520.0, "step": 7114 }, { "epoch": 0.9051011321714795, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.193442344665527, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8577890396118164, "num_tokens": 271445138.0, "step": 7115 }, { "epoch": 0.90522834245007, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.166293144226074, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8686428666114807, "num_tokens": 271479147.0, "step": 7116 }, { "epoch": 0.9053555527286605, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.212690353393555, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.865754246711731, "num_tokens": 271512477.0, "step": 7117 }, { "epoch": 0.905482763007251, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.15041446685791, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.850285530090332, "num_tokens": 271552355.0, "step": 7118 }, { "epoch": 0.9056099732858415, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.19094181060791, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8601435422897339, "num_tokens": 271595825.0, "step": 7119 }, { "epoch": 0.905737183564432, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.142095565795898, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8511534333229065, "num_tokens": 271627384.0, "step": 7120 }, { "epoch": 0.9058643938430225, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.220548629760742, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.872539222240448, "num_tokens": 271666711.0, "step": 7121 }, { "epoch": 0.9059916041216131, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.214070320129395, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8692480325698853, "num_tokens": 271703145.0, "step": 7122 }, { "epoch": 0.9061188144002036, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.125646591186523, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8517078161239624, "num_tokens": 271743131.0, "step": 7123 }, { "epoch": 0.906246024678794, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.252562522888184, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.867409348487854, "num_tokens": 271778245.0, "step": 7124 }, { "epoch": 0.9063732349573845, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.165107727050781, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8542917370796204, "num_tokens": 271817279.0, "step": 7125 }, { "epoch": 0.9065004452359751, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.176934242248535, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8711918592453003, "num_tokens": 271857136.0, "step": 7126 }, { "epoch": 0.9066276555145656, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.129525184631348, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8724545240402222, "num_tokens": 271891651.0, "step": 7127 }, { "epoch": 0.9067548657931561, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.171780586242676, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8687021136283875, "num_tokens": 271925805.0, "step": 7128 }, { "epoch": 0.9068820760717466, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.126290321350098, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8682825565338135, "num_tokens": 271964954.0, "step": 7129 }, { "epoch": 0.9070092863503371, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.145519256591797, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8626347184181213, "num_tokens": 272000331.0, "step": 7130 }, { "epoch": 0.9071364966289276, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.136878967285156, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8706942796707153, "num_tokens": 272037853.0, "step": 7131 }, { "epoch": 0.9072637069075181, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.107232093811035, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.873481273651123, "num_tokens": 272077663.0, "step": 7132 }, { "epoch": 0.9073909171861086, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.15477180480957, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8596247434616089, "num_tokens": 272116942.0, "step": 7133 }, { "epoch": 0.9075181274646992, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.163495063781738, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.856249213218689, "num_tokens": 272151319.0, "step": 7134 }, { "epoch": 0.9076453377432897, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.134953498840332, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8512536883354187, "num_tokens": 272190817.0, "step": 7135 }, { "epoch": 0.9077725480218801, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.139713287353516, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8572864532470703, "num_tokens": 272230461.0, "step": 7136 }, { "epoch": 0.9078997583004706, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.19983196258545, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.85762619972229, "num_tokens": 272268510.0, "step": 7137 }, { "epoch": 0.9080269685790612, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.165068626403809, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8646365404129028, "num_tokens": 272312430.0, "step": 7138 }, { "epoch": 0.9081541788576517, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.139471054077148, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8622024655342102, "num_tokens": 272348437.0, "step": 7139 }, { "epoch": 0.9082813891362422, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.168449401855469, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8570938110351562, "num_tokens": 272378902.0, "step": 7140 }, { "epoch": 0.9084085994148328, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189825057983398, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8769727945327759, "num_tokens": 272421257.0, "step": 7141 }, { "epoch": 0.9085358096934232, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.110885620117188, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.862565279006958, "num_tokens": 272459800.0, "step": 7142 }, { "epoch": 0.9086630199720137, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.128028869628906, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8818380832672119, "num_tokens": 272501245.0, "step": 7143 }, { "epoch": 0.9087902302506042, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.198301315307617, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8644297122955322, "num_tokens": 272535873.0, "step": 7144 }, { "epoch": 0.9089174405291948, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.14429759979248, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8614643812179565, "num_tokens": 272575213.0, "step": 7145 }, { "epoch": 0.9090446508077853, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.121137619018555, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8653836250305176, "num_tokens": 272611960.0, "step": 7146 }, { "epoch": 0.9091718610863758, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.116552352905273, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8638370633125305, "num_tokens": 272650895.0, "step": 7147 }, { "epoch": 0.9092990713649662, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.165085792541504, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8644318580627441, "num_tokens": 272690617.0, "step": 7148 }, { "epoch": 0.9094262816435568, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.150201797485352, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8554131984710693, "num_tokens": 272732526.0, "step": 7149 }, { "epoch": 0.9095534919221473, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.161321640014648, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8617198467254639, "num_tokens": 272769882.0, "step": 7150 }, { "epoch": 0.9096807022007378, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.170416831970215, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.866631269454956, "num_tokens": 272809614.0, "step": 7151 }, { "epoch": 0.9098079124793284, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.1531400680542, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8565953373908997, "num_tokens": 272844819.0, "step": 7152 }, { "epoch": 0.9099351227579189, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.225117683410645, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8690637946128845, "num_tokens": 272882798.0, "step": 7153 }, { "epoch": 0.9100623330365093, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.198701858520508, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8665753602981567, "num_tokens": 272917472.0, "step": 7154 }, { "epoch": 0.9101895433150998, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.197352409362793, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8527736067771912, "num_tokens": 272959947.0, "step": 7155 }, { "epoch": 0.9103167535936904, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.150805473327637, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8464792966842651, "num_tokens": 272998194.0, "step": 7156 }, { "epoch": 0.9104439638722809, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.221573829650879, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8811073899269104, "num_tokens": 273034688.0, "step": 7157 }, { "epoch": 0.9105711741508714, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.201125144958496, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.874264657497406, "num_tokens": 273071575.0, "step": 7158 }, { "epoch": 0.9106983844294619, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.17050552368164, "learning_rate": 1e-06, "loss": 0.507, "mean_token_accuracy": 0.8377441763877869, "num_tokens": 273106106.0, "step": 7159 }, { "epoch": 0.9108255947080524, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.20682430267334, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8579950332641602, "num_tokens": 273142959.0, "step": 7160 }, { "epoch": 0.9109528049866429, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.184206008911133, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8484175205230713, "num_tokens": 273182781.0, "step": 7161 }, { "epoch": 0.9110800152652334, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.168573379516602, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8697659969329834, "num_tokens": 273225738.0, "step": 7162 }, { "epoch": 0.9112072255438239, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.114119529724121, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8707306385040283, "num_tokens": 273269166.0, "step": 7163 }, { "epoch": 0.9113344358224145, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.22925090789795, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8584014177322388, "num_tokens": 273309093.0, "step": 7164 }, { "epoch": 0.911461646101005, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.196334838867188, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8685228824615479, "num_tokens": 273346857.0, "step": 7165 }, { "epoch": 0.9115888563795955, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.18820858001709, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8588498830795288, "num_tokens": 273385720.0, "step": 7166 }, { "epoch": 0.9117160666581859, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.172679901123047, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.864364504814148, "num_tokens": 273431079.0, "step": 7167 }, { "epoch": 0.9118432769367765, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.1463041305542, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8644298315048218, "num_tokens": 273477316.0, "step": 7168 }, { "epoch": 0.911970487215367, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.266551971435547, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8676923513412476, "num_tokens": 273512831.0, "step": 7169 }, { "epoch": 0.9120976974939575, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.131747245788574, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8571141958236694, "num_tokens": 273551927.0, "step": 7170 }, { "epoch": 0.912224907772548, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.156192779541016, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8667988777160645, "num_tokens": 273583053.0, "step": 7171 }, { "epoch": 0.9123521180511386, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.183089256286621, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8537025451660156, "num_tokens": 273617953.0, "step": 7172 }, { "epoch": 0.912479328329729, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.176609992980957, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.850135862827301, "num_tokens": 273649959.0, "step": 7173 }, { "epoch": 0.9126065386083195, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.199926376342773, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8536294102668762, "num_tokens": 273682666.0, "step": 7174 }, { "epoch": 0.9127337488869101, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.166787147521973, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8823066353797913, "num_tokens": 273718641.0, "step": 7175 }, { "epoch": 0.9128609591655006, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.13543701171875, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8729152083396912, "num_tokens": 273760502.0, "step": 7176 }, { "epoch": 0.9129881694440911, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.227171897888184, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8576870560646057, "num_tokens": 273795068.0, "step": 7177 }, { "epoch": 0.9131153797226816, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.178033828735352, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8455110788345337, "num_tokens": 273825099.0, "step": 7178 }, { "epoch": 0.9132425900012721, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.147599220275879, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8695286512374878, "num_tokens": 273862566.0, "step": 7179 }, { "epoch": 0.9133698002798626, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.174579620361328, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8644188642501831, "num_tokens": 273896107.0, "step": 7180 }, { "epoch": 0.9134970105584531, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.094037055969238, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8684963583946228, "num_tokens": 273934066.0, "step": 7181 }, { "epoch": 0.9136242208370436, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.13509464263916, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8625568151473999, "num_tokens": 273971843.0, "step": 7182 }, { "epoch": 0.9137514311156342, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.208303451538086, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8562244176864624, "num_tokens": 274001646.0, "step": 7183 }, { "epoch": 0.9138786413942247, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.14489459991455, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8486343026161194, "num_tokens": 274042656.0, "step": 7184 }, { "epoch": 0.9140058516728151, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.124521255493164, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8498173356056213, "num_tokens": 274084609.0, "step": 7185 }, { "epoch": 0.9141330619514056, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.192500114440918, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.866904616355896, "num_tokens": 274127503.0, "step": 7186 }, { "epoch": 0.9142602722299962, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.115740776062012, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8675526976585388, "num_tokens": 274164126.0, "step": 7187 }, { "epoch": 0.9143874825085867, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.222122192382812, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8631706237792969, "num_tokens": 274198104.0, "step": 7188 }, { "epoch": 0.9145146927871772, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.22122573852539, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8529525995254517, "num_tokens": 274233487.0, "step": 7189 }, { "epoch": 0.9146419030657678, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.136804580688477, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8681807518005371, "num_tokens": 274267317.0, "step": 7190 }, { "epoch": 0.9147691133443582, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.143518447875977, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8568956851959229, "num_tokens": 274304169.0, "step": 7191 }, { "epoch": 0.9148963236229487, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.197009086608887, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.863771378993988, "num_tokens": 274343265.0, "step": 7192 }, { "epoch": 0.9150235339015392, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.172040939331055, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8474898934364319, "num_tokens": 274380756.0, "step": 7193 }, { "epoch": 0.9151507441801298, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.151825904846191, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8730936050415039, "num_tokens": 274422035.0, "step": 7194 }, { "epoch": 0.9152779544587203, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.159908294677734, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.853513240814209, "num_tokens": 274460875.0, "step": 7195 }, { "epoch": 0.9154051647373108, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.260273933410645, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8660684823989868, "num_tokens": 274495807.0, "step": 7196 }, { "epoch": 0.9155323750159012, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.208030700683594, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.870975911617279, "num_tokens": 274533098.0, "step": 7197 }, { "epoch": 0.9156595852944918, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.151579856872559, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8589729070663452, "num_tokens": 274565282.0, "step": 7198 }, { "epoch": 0.9157867955730823, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.196293830871582, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8671822547912598, "num_tokens": 274605401.0, "step": 7199 }, { "epoch": 0.9159140058516728, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189525604248047, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8574683666229248, "num_tokens": 274645018.0, "step": 7200 }, { "epoch": 0.9160412161302633, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.194496154785156, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8726320862770081, "num_tokens": 274681541.0, "step": 7201 }, { "epoch": 0.9161684264088539, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.217169761657715, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8674774169921875, "num_tokens": 274716883.0, "step": 7202 }, { "epoch": 0.9162956366874443, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.190950393676758, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8577888011932373, "num_tokens": 274753038.0, "step": 7203 }, { "epoch": 0.9164228469660348, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.220807075500488, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8700866103172302, "num_tokens": 274790760.0, "step": 7204 }, { "epoch": 0.9165500572446253, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.136665344238281, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8559167385101318, "num_tokens": 274830872.0, "step": 7205 }, { "epoch": 0.9166772675232159, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.230749130249023, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8508203625679016, "num_tokens": 274869832.0, "step": 7206 }, { "epoch": 0.9168044778018064, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.138143539428711, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8558290004730225, "num_tokens": 274903853.0, "step": 7207 }, { "epoch": 0.9169316880803969, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.218650817871094, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8625081777572632, "num_tokens": 274942911.0, "step": 7208 }, { "epoch": 0.9170588983589874, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.12392520904541, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8705084919929504, "num_tokens": 274986427.0, "step": 7209 }, { "epoch": 0.9171861086375779, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189976692199707, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8656857013702393, "num_tokens": 275021557.0, "step": 7210 }, { "epoch": 0.9173133189161684, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.197599411010742, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8520025610923767, "num_tokens": 275061308.0, "step": 7211 }, { "epoch": 0.9174405291947589, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.113728523254395, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8556855320930481, "num_tokens": 275101497.0, "step": 7212 }, { "epoch": 0.9175677394733495, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.212348937988281, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8521727323532104, "num_tokens": 275138317.0, "step": 7213 }, { "epoch": 0.91769494975194, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.135689735412598, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8623090982437134, "num_tokens": 275180119.0, "step": 7214 }, { "epoch": 0.9178221600305305, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.213911056518555, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8612246513366699, "num_tokens": 275212083.0, "step": 7215 }, { "epoch": 0.9179493703091209, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.222260475158691, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8534310460090637, "num_tokens": 275247872.0, "step": 7216 }, { "epoch": 0.9180765805877115, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.14080810546875, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8673760890960693, "num_tokens": 275287022.0, "step": 7217 }, { "epoch": 0.918203790866302, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.244522094726562, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8650555610656738, "num_tokens": 275322259.0, "step": 7218 }, { "epoch": 0.9183310011448925, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.222747802734375, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8768644332885742, "num_tokens": 275357410.0, "step": 7219 }, { "epoch": 0.918458211423483, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.210053443908691, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8754698038101196, "num_tokens": 275388720.0, "step": 7220 }, { "epoch": 0.9185854217020736, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.191417694091797, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8692848682403564, "num_tokens": 275427699.0, "step": 7221 }, { "epoch": 0.918712631980664, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.113627433776855, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8659012317657471, "num_tokens": 275472220.0, "step": 7222 }, { "epoch": 0.9188398422592545, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.192424774169922, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8512201905250549, "num_tokens": 275506693.0, "step": 7223 }, { "epoch": 0.918967052537845, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189826011657715, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.86630779504776, "num_tokens": 275547295.0, "step": 7224 }, { "epoch": 0.9190942628164356, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.180973052978516, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8591774702072144, "num_tokens": 275591552.0, "step": 7225 }, { "epoch": 0.9192214730950261, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.205154418945312, "learning_rate": 1e-06, "loss": 0.5067, "mean_token_accuracy": 0.8378541469573975, "num_tokens": 275632197.0, "step": 7226 }, { "epoch": 0.9193486833736166, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.201142311096191, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8562465310096741, "num_tokens": 275670680.0, "step": 7227 }, { "epoch": 0.919475893652207, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189581871032715, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.861198365688324, "num_tokens": 275709315.0, "step": 7228 }, { "epoch": 0.9196031039307976, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.228165626525879, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8621736764907837, "num_tokens": 275740405.0, "step": 7229 }, { "epoch": 0.9197303142093881, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.160299301147461, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.880043625831604, "num_tokens": 275780915.0, "step": 7230 }, { "epoch": 0.9198575244879786, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.248483657836914, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8579130172729492, "num_tokens": 275817832.0, "step": 7231 }, { "epoch": 0.9199847347665692, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.11867904663086, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.866047739982605, "num_tokens": 275857621.0, "step": 7232 }, { "epoch": 0.9201119450451597, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.214337348937988, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8593292236328125, "num_tokens": 275898917.0, "step": 7233 }, { "epoch": 0.9202391553237501, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.221250534057617, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8675472736358643, "num_tokens": 275930151.0, "step": 7234 }, { "epoch": 0.9203663656023406, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.182376861572266, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8693549036979675, "num_tokens": 275966901.0, "step": 7235 }, { "epoch": 0.9204935758809312, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.204423904418945, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.884581446647644, "num_tokens": 276002845.0, "step": 7236 }, { "epoch": 0.9206207861595217, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.155890464782715, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8612351417541504, "num_tokens": 276038476.0, "step": 7237 }, { "epoch": 0.9207479964381122, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.247562408447266, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8583750128746033, "num_tokens": 276077867.0, "step": 7238 }, { "epoch": 0.9208752067167028, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.139081954956055, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8528358340263367, "num_tokens": 276123289.0, "step": 7239 }, { "epoch": 0.9210024169952932, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.18751335144043, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8507175445556641, "num_tokens": 276157357.0, "step": 7240 }, { "epoch": 0.9211296272738837, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.155452728271484, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8834895491600037, "num_tokens": 276195548.0, "step": 7241 }, { "epoch": 0.9212568375524742, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.223386764526367, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8656377196311951, "num_tokens": 276233533.0, "step": 7242 }, { "epoch": 0.9213840478310648, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.224129676818848, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8663730621337891, "num_tokens": 276274310.0, "step": 7243 }, { "epoch": 0.9215112581096553, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.19306755065918, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8602280616760254, "num_tokens": 276310272.0, "step": 7244 }, { "epoch": 0.9216384683882458, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.232975006103516, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8569298982620239, "num_tokens": 276350334.0, "step": 7245 }, { "epoch": 0.9217656786668362, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.209563255310059, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8579453229904175, "num_tokens": 276393913.0, "step": 7246 }, { "epoch": 0.9218928889454268, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.224980354309082, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8586113452911377, "num_tokens": 276430368.0, "step": 7247 }, { "epoch": 0.9220200992240173, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.231505393981934, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8567743301391602, "num_tokens": 276470930.0, "step": 7248 }, { "epoch": 0.9221473095026078, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.175039291381836, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8674991130828857, "num_tokens": 276509359.0, "step": 7249 }, { "epoch": 0.9222745197811983, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.212204933166504, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8542523384094238, "num_tokens": 276546980.0, "step": 7250 }, { "epoch": 0.9224017300597889, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.160841941833496, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.850311279296875, "num_tokens": 276586212.0, "step": 7251 }, { "epoch": 0.9225289403383793, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.238051414489746, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8637397885322571, "num_tokens": 276623946.0, "step": 7252 }, { "epoch": 0.9226561506169698, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.1254301071167, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8574848175048828, "num_tokens": 276667567.0, "step": 7253 }, { "epoch": 0.9227833608955603, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.235843658447266, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8564474582672119, "num_tokens": 276708479.0, "step": 7254 }, { "epoch": 0.9229105711741509, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.17442798614502, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8566085696220398, "num_tokens": 276744627.0, "step": 7255 }, { "epoch": 0.9230377814527414, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.277982711791992, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8635351657867432, "num_tokens": 276782830.0, "step": 7256 }, { "epoch": 0.9231649917313319, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.187670707702637, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8607386946678162, "num_tokens": 276825745.0, "step": 7257 }, { "epoch": 0.9232922020099223, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.245976448059082, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8680901527404785, "num_tokens": 276866743.0, "step": 7258 }, { "epoch": 0.9234194122885129, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.192416191101074, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.858188271522522, "num_tokens": 276903884.0, "step": 7259 }, { "epoch": 0.9235466225671034, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.170738220214844, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8688738346099854, "num_tokens": 276939142.0, "step": 7260 }, { "epoch": 0.9236738328456939, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.258110046386719, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8637846112251282, "num_tokens": 276979041.0, "step": 7261 }, { "epoch": 0.9238010431242845, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.175943374633789, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8743619918823242, "num_tokens": 277015590.0, "step": 7262 }, { "epoch": 0.923928253402875, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.1417875289917, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.855188250541687, "num_tokens": 277055745.0, "step": 7263 }, { "epoch": 0.9240554636814655, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.261724472045898, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8583102822303772, "num_tokens": 277098856.0, "step": 7264 }, { "epoch": 0.9241826739600559, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.07132625579834, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8733196258544922, "num_tokens": 277139165.0, "step": 7265 }, { "epoch": 0.9243098842386465, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.233925819396973, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8676255345344543, "num_tokens": 277171879.0, "step": 7266 }, { "epoch": 0.924437094517237, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.168598175048828, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8634951114654541, "num_tokens": 277211719.0, "step": 7267 }, { "epoch": 0.9245643047958275, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.161620140075684, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8543614149093628, "num_tokens": 277252675.0, "step": 7268 }, { "epoch": 0.924691515074418, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.245706558227539, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8510428667068481, "num_tokens": 277287529.0, "step": 7269 }, { "epoch": 0.9248187253530086, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.180177688598633, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8618085980415344, "num_tokens": 277328392.0, "step": 7270 }, { "epoch": 0.924945935631599, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.204800605773926, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8718975186347961, "num_tokens": 277368412.0, "step": 7271 }, { "epoch": 0.9250731459101895, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.218910217285156, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8674048185348511, "num_tokens": 277404373.0, "step": 7272 }, { "epoch": 0.92520035618878, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.21772575378418, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8595338463783264, "num_tokens": 277444988.0, "step": 7273 }, { "epoch": 0.9253275664673706, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.186964988708496, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8715177774429321, "num_tokens": 277487501.0, "step": 7274 }, { "epoch": 0.9254547767459611, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.220870971679688, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8599472045898438, "num_tokens": 277529056.0, "step": 7275 }, { "epoch": 0.9255819870245516, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.235072135925293, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8754555583000183, "num_tokens": 277563022.0, "step": 7276 }, { "epoch": 0.925709197303142, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.16092586517334, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8693090081214905, "num_tokens": 277603879.0, "step": 7277 }, { "epoch": 0.9258364075817326, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.233551979064941, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8583455085754395, "num_tokens": 277643711.0, "step": 7278 }, { "epoch": 0.9259636178603231, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.236031532287598, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8620457053184509, "num_tokens": 277680667.0, "step": 7279 }, { "epoch": 0.9260908281389136, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.238543510437012, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8611501455307007, "num_tokens": 277715548.0, "step": 7280 }, { "epoch": 0.9262180384175042, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.21414852142334, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8619239330291748, "num_tokens": 277745287.0, "step": 7281 }, { "epoch": 0.9263452486960947, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.211358070373535, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8562458157539368, "num_tokens": 277781484.0, "step": 7282 }, { "epoch": 0.9264724589746851, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.129344940185547, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.865172266960144, "num_tokens": 277818429.0, "step": 7283 }, { "epoch": 0.9265996692532756, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24594497680664, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8701208829879761, "num_tokens": 277853077.0, "step": 7284 }, { "epoch": 0.9267268795318662, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.182058334350586, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8804558515548706, "num_tokens": 277893997.0, "step": 7285 }, { "epoch": 0.9268540898104567, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.185853958129883, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8519659638404846, "num_tokens": 277934680.0, "step": 7286 }, { "epoch": 0.9269813000890472, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.192225456237793, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8555234670639038, "num_tokens": 277980370.0, "step": 7287 }, { "epoch": 0.9271085103676378, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.242806434631348, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8740253448486328, "num_tokens": 278017257.0, "step": 7288 }, { "epoch": 0.9272357206462282, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.234349250793457, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8627169728279114, "num_tokens": 278054835.0, "step": 7289 }, { "epoch": 0.9273629309248187, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.223795890808105, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8607944250106812, "num_tokens": 278089510.0, "step": 7290 }, { "epoch": 0.9274901412034092, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189156532287598, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8645622730255127, "num_tokens": 278126179.0, "step": 7291 }, { "epoch": 0.9276173514819998, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.245667457580566, "learning_rate": 1e-06, "loss": 0.4989, "mean_token_accuracy": 0.840717077255249, "num_tokens": 278170837.0, "step": 7292 }, { "epoch": 0.9277445617605903, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.213630676269531, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8672877550125122, "num_tokens": 278207410.0, "step": 7293 }, { "epoch": 0.9278717720391808, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.252917289733887, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8829208612442017, "num_tokens": 278240495.0, "step": 7294 }, { "epoch": 0.9279989823177712, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.171525955200195, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8541459441184998, "num_tokens": 278278449.0, "step": 7295 }, { "epoch": 0.9281261925963618, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.211624145507812, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8720947504043579, "num_tokens": 278321434.0, "step": 7296 }, { "epoch": 0.9282534028749523, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.221169471740723, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8582016229629517, "num_tokens": 278363730.0, "step": 7297 }, { "epoch": 0.9283806131535428, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.289690971374512, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8640376329421997, "num_tokens": 278395493.0, "step": 7298 }, { "epoch": 0.9285078234321333, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.204652786254883, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8635923862457275, "num_tokens": 278433024.0, "step": 7299 }, { "epoch": 0.9286350337107239, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.251029014587402, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8642780780792236, "num_tokens": 278471311.0, "step": 7300 }, { "epoch": 0.9287622439893143, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.1940279006958, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8823356628417969, "num_tokens": 278509151.0, "step": 7301 }, { "epoch": 0.9288894542679048, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.25161361694336, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8697922229766846, "num_tokens": 278548743.0, "step": 7302 }, { "epoch": 0.9290166645464953, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.190864562988281, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.873999297618866, "num_tokens": 278590722.0, "step": 7303 }, { "epoch": 0.9291438748250859, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.231731414794922, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.8503394722938538, "num_tokens": 278628150.0, "step": 7304 }, { "epoch": 0.9292710851036764, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.219598770141602, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8468808531761169, "num_tokens": 278672543.0, "step": 7305 }, { "epoch": 0.9293982953822669, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.19786262512207, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.872035562992096, "num_tokens": 278707234.0, "step": 7306 }, { "epoch": 0.9295255056608573, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.248878479003906, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8654285669326782, "num_tokens": 278734678.0, "step": 7307 }, { "epoch": 0.9296527159394479, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.224176406860352, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8729512691497803, "num_tokens": 278778284.0, "step": 7308 }, { "epoch": 0.9297799262180384, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.208023071289062, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.862673282623291, "num_tokens": 278815363.0, "step": 7309 }, { "epoch": 0.9299071364966289, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.288718223571777, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8506202101707458, "num_tokens": 278860777.0, "step": 7310 }, { "epoch": 0.9300343467752195, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.199418067932129, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8773602247238159, "num_tokens": 278890552.0, "step": 7311 }, { "epoch": 0.93016155705381, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.358222007751465, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8528372049331665, "num_tokens": 278924741.0, "step": 7312 }, { "epoch": 0.9302887673324005, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.135682106018066, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8721300363540649, "num_tokens": 278965185.0, "step": 7313 }, { "epoch": 0.9304159776109909, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.295506477355957, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8523851633071899, "num_tokens": 278998961.0, "step": 7314 }, { "epoch": 0.9305431878895815, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.257698059082031, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8501882553100586, "num_tokens": 279032689.0, "step": 7315 }, { "epoch": 0.930670398168172, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.21372127532959, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.849453866481781, "num_tokens": 279080305.0, "step": 7316 }, { "epoch": 0.9307976084467625, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.247248649597168, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8639200925827026, "num_tokens": 279120581.0, "step": 7317 }, { "epoch": 0.930924818725353, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.203177452087402, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.875461220741272, "num_tokens": 279165050.0, "step": 7318 }, { "epoch": 0.9310520290039436, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.195944786071777, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8671590089797974, "num_tokens": 279209393.0, "step": 7319 }, { "epoch": 0.931179239282534, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.268624305725098, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8507614135742188, "num_tokens": 279250569.0, "step": 7320 }, { "epoch": 0.9313064495611245, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.244495391845703, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8694788217544556, "num_tokens": 279291154.0, "step": 7321 }, { "epoch": 0.931433659839715, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.242496490478516, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8623316287994385, "num_tokens": 279326967.0, "step": 7322 }, { "epoch": 0.9315608701183056, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.241716384887695, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8751088380813599, "num_tokens": 279367129.0, "step": 7323 }, { "epoch": 0.9316880803968961, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.268417358398438, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8693444728851318, "num_tokens": 279403651.0, "step": 7324 }, { "epoch": 0.9318152906754866, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.314379692077637, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8557252883911133, "num_tokens": 279439166.0, "step": 7325 }, { "epoch": 0.931942500954077, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.230621337890625, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8617477416992188, "num_tokens": 279473489.0, "step": 7326 }, { "epoch": 0.9320697112326676, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.268050193786621, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8670659065246582, "num_tokens": 279511709.0, "step": 7327 }, { "epoch": 0.9321969215112581, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.229323387145996, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8504706025123596, "num_tokens": 279551638.0, "step": 7328 }, { "epoch": 0.9323241317898486, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.266825675964355, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8750468492507935, "num_tokens": 279587304.0, "step": 7329 }, { "epoch": 0.9324513420684392, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.28725814819336, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8548913598060608, "num_tokens": 279628138.0, "step": 7330 }, { "epoch": 0.9325785523470297, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.216180801391602, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8680524826049805, "num_tokens": 279663064.0, "step": 7331 }, { "epoch": 0.9327057626256201, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.208657264709473, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8552253246307373, "num_tokens": 279704937.0, "step": 7332 }, { "epoch": 0.9328329729042106, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.270938873291016, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8652232885360718, "num_tokens": 279738215.0, "step": 7333 }, { "epoch": 0.9329601831828012, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.27088737487793, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8472183346748352, "num_tokens": 279771996.0, "step": 7334 }, { "epoch": 0.9330873934613917, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.259115219116211, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8772361278533936, "num_tokens": 279810037.0, "step": 7335 }, { "epoch": 0.9332146037399822, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24906063079834, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8575789928436279, "num_tokens": 279847394.0, "step": 7336 }, { "epoch": 0.9333418140185727, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.276427268981934, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8491533994674683, "num_tokens": 279879057.0, "step": 7337 }, { "epoch": 0.9334690242971632, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.25743293762207, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8573818802833557, "num_tokens": 279911775.0, "step": 7338 }, { "epoch": 0.9335962345757537, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.21924877166748, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8655603528022766, "num_tokens": 279947661.0, "step": 7339 }, { "epoch": 0.9337234448543442, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.259472846984863, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8643425107002258, "num_tokens": 279977507.0, "step": 7340 }, { "epoch": 0.9338506551329347, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.258430480957031, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.859636664390564, "num_tokens": 280016749.0, "step": 7341 }, { "epoch": 0.9339778654115253, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.277242660522461, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8652909994125366, "num_tokens": 280056092.0, "step": 7342 }, { "epoch": 0.9341050756901158, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.266585350036621, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.875685453414917, "num_tokens": 280089803.0, "step": 7343 }, { "epoch": 0.9342322859687062, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.281603813171387, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8516605496406555, "num_tokens": 280126190.0, "step": 7344 }, { "epoch": 0.9343594962472968, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.223311424255371, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8742862343788147, "num_tokens": 280162743.0, "step": 7345 }, { "epoch": 0.9344867065258873, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.268075942993164, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8627326488494873, "num_tokens": 280203676.0, "step": 7346 }, { "epoch": 0.9346139168044778, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.236650466918945, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8794431686401367, "num_tokens": 280241765.0, "step": 7347 }, { "epoch": 0.9347411270830683, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.340201377868652, "learning_rate": 1e-06, "loss": 0.5221, "mean_token_accuracy": 0.8426566123962402, "num_tokens": 280280103.0, "step": 7348 }, { "epoch": 0.9348683373616589, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.182592391967773, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8471693992614746, "num_tokens": 280314674.0, "step": 7349 }, { "epoch": 0.9349955476402493, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.303274154663086, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.864532470703125, "num_tokens": 280353890.0, "step": 7350 }, { "epoch": 0.9351227579188398, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.185247421264648, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8561042547225952, "num_tokens": 280396403.0, "step": 7351 }, { "epoch": 0.9352499681974303, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.286866188049316, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8631503582000732, "num_tokens": 280436154.0, "step": 7352 }, { "epoch": 0.9353771784760209, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.237041473388672, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8706275224685669, "num_tokens": 280476480.0, "step": 7353 }, { "epoch": 0.9355043887546114, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.293354034423828, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8666635751724243, "num_tokens": 280505976.0, "step": 7354 }, { "epoch": 0.9356315990332019, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.195594787597656, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8590120673179626, "num_tokens": 280548365.0, "step": 7355 }, { "epoch": 0.9357588093117923, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.21467399597168, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8813979625701904, "num_tokens": 280588074.0, "step": 7356 }, { "epoch": 0.9358860195903829, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.26052474975586, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8722756505012512, "num_tokens": 280626915.0, "step": 7357 }, { "epoch": 0.9360132298689734, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.292771339416504, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8577675819396973, "num_tokens": 280660088.0, "step": 7358 }, { "epoch": 0.9361404401475639, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.27649974822998, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8613161444664001, "num_tokens": 280701326.0, "step": 7359 }, { "epoch": 0.9362676504261545, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.228307723999023, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8728936314582825, "num_tokens": 280740256.0, "step": 7360 }, { "epoch": 0.936394860704745, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.293805122375488, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8766447305679321, "num_tokens": 280773515.0, "step": 7361 }, { "epoch": 0.9365220709833355, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.29002571105957, "learning_rate": 1e-06, "loss": 0.4987, "mean_token_accuracy": 0.8378912210464478, "num_tokens": 280817207.0, "step": 7362 }, { "epoch": 0.9366492812619259, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.255335807800293, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8503545522689819, "num_tokens": 280860253.0, "step": 7363 }, { "epoch": 0.9367764915405165, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.285938262939453, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.859993577003479, "num_tokens": 280892634.0, "step": 7364 }, { "epoch": 0.936903701819107, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.246923446655273, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8717612624168396, "num_tokens": 280930912.0, "step": 7365 }, { "epoch": 0.9370309120976975, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.17304801940918, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8672510385513306, "num_tokens": 280971588.0, "step": 7366 }, { "epoch": 0.937158122376288, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.322785377502441, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8706265687942505, "num_tokens": 281009898.0, "step": 7367 }, { "epoch": 0.9372853326548786, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.255995750427246, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8554872870445251, "num_tokens": 281046726.0, "step": 7368 }, { "epoch": 0.937412542933469, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.240474700927734, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8773399591445923, "num_tokens": 281085163.0, "step": 7369 }, { "epoch": 0.9375397532120595, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.245808601379395, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8678895235061646, "num_tokens": 281128660.0, "step": 7370 }, { "epoch": 0.93766696349065, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.288713455200195, "learning_rate": 1e-06, "loss": 0.4943, "mean_token_accuracy": 0.84539395570755, "num_tokens": 281163780.0, "step": 7371 }, { "epoch": 0.9377941737692406, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.255338668823242, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8700974583625793, "num_tokens": 281204247.0, "step": 7372 }, { "epoch": 0.9379213840478311, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.18208122253418, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8878498077392578, "num_tokens": 281239518.0, "step": 7373 }, { "epoch": 0.9380485943264216, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.277432441711426, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8721885681152344, "num_tokens": 281275916.0, "step": 7374 }, { "epoch": 0.938175804605012, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.252410888671875, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8738248348236084, "num_tokens": 281307615.0, "step": 7375 }, { "epoch": 0.9383030148836026, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.261667251586914, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.868677020072937, "num_tokens": 281342677.0, "step": 7376 }, { "epoch": 0.9384302251621931, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.269855499267578, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8594136238098145, "num_tokens": 281379545.0, "step": 7377 }, { "epoch": 0.9385574354407836, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.230074882507324, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8449140787124634, "num_tokens": 281424169.0, "step": 7378 }, { "epoch": 0.9386846457193742, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.214386940002441, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8840848207473755, "num_tokens": 281461306.0, "step": 7379 }, { "epoch": 0.9388118559979647, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24044418334961, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8767584562301636, "num_tokens": 281498753.0, "step": 7380 }, { "epoch": 0.9389390662765551, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.274688720703125, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8729560375213623, "num_tokens": 281537163.0, "step": 7381 }, { "epoch": 0.9390662765551456, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.285158157348633, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8554871678352356, "num_tokens": 281577926.0, "step": 7382 }, { "epoch": 0.9391934868337362, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.261144638061523, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8830748796463013, "num_tokens": 281621283.0, "step": 7383 }, { "epoch": 0.9393206971123267, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.308494567871094, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8669841885566711, "num_tokens": 281658336.0, "step": 7384 }, { "epoch": 0.9394479073909172, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.29340934753418, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8670037984848022, "num_tokens": 281695137.0, "step": 7385 }, { "epoch": 0.9395751176695077, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.273961067199707, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8779785633087158, "num_tokens": 281733570.0, "step": 7386 }, { "epoch": 0.9397023279480982, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.292499542236328, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8649742603302002, "num_tokens": 281771163.0, "step": 7387 }, { "epoch": 0.9398295382266887, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.268562316894531, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8672153949737549, "num_tokens": 281813715.0, "step": 7388 }, { "epoch": 0.9399567485052792, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.186514854431152, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8621975183486938, "num_tokens": 281856921.0, "step": 7389 }, { "epoch": 0.9400839587838697, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.321147918701172, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8550100922584534, "num_tokens": 281893273.0, "step": 7390 }, { "epoch": 0.9402111690624603, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24245548248291, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8689011335372925, "num_tokens": 281927452.0, "step": 7391 }, { "epoch": 0.9403383793410508, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24582576751709, "learning_rate": 1e-06, "loss": 0.3373, "mean_token_accuracy": 0.888745129108429, "num_tokens": 281965359.0, "step": 7392 }, { "epoch": 0.9404655896196412, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.240448951721191, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.861700177192688, "num_tokens": 281999209.0, "step": 7393 }, { "epoch": 0.9405927998982317, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.278383255004883, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8543392419815063, "num_tokens": 282042129.0, "step": 7394 }, { "epoch": 0.9407200101768223, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.253480911254883, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8551927804946899, "num_tokens": 282087941.0, "step": 7395 }, { "epoch": 0.9408472204554128, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.310080528259277, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8688649535179138, "num_tokens": 282123130.0, "step": 7396 }, { "epoch": 0.9409744307340033, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.340574264526367, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8600807189941406, "num_tokens": 282161142.0, "step": 7397 }, { "epoch": 0.9411016410125939, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.309406280517578, "learning_rate": 1e-06, "loss": 0.3416, "mean_token_accuracy": 0.888068675994873, "num_tokens": 282196363.0, "step": 7398 }, { "epoch": 0.9412288512911843, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.273995399475098, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8450914025306702, "num_tokens": 282233202.0, "step": 7399 }, { "epoch": 0.9413560615697748, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.32758903503418, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8657006025314331, "num_tokens": 282270976.0, "step": 7400 }, { "epoch": 0.9414832718483653, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.248832702636719, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8742576837539673, "num_tokens": 282312223.0, "step": 7401 }, { "epoch": 0.9416104821269559, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.274958610534668, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.848455011844635, "num_tokens": 282350418.0, "step": 7402 }, { "epoch": 0.9417376924055464, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.303610801696777, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8650876879692078, "num_tokens": 282380751.0, "step": 7403 }, { "epoch": 0.9418649026841369, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.282037734985352, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8764384388923645, "num_tokens": 282420248.0, "step": 7404 }, { "epoch": 0.9419921129627273, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.259114265441895, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8687292337417603, "num_tokens": 282457951.0, "step": 7405 }, { "epoch": 0.9421193232413179, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.266216278076172, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8634214401245117, "num_tokens": 282507379.0, "step": 7406 }, { "epoch": 0.9422465335199084, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.191458702087402, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8525290489196777, "num_tokens": 282546905.0, "step": 7407 }, { "epoch": 0.9423737437984989, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.259833335876465, "learning_rate": 1e-06, "loss": 0.5169, "mean_token_accuracy": 0.8359731435775757, "num_tokens": 282591368.0, "step": 7408 }, { "epoch": 0.9425009540770894, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.294526100158691, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8614758849143982, "num_tokens": 282617331.0, "step": 7409 }, { "epoch": 0.94262816435568, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.223718643188477, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8772704005241394, "num_tokens": 282652346.0, "step": 7410 }, { "epoch": 0.9427553746342705, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.318212509155273, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8679136037826538, "num_tokens": 282691564.0, "step": 7411 }, { "epoch": 0.9428825849128609, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.305462837219238, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8742512464523315, "num_tokens": 282728186.0, "step": 7412 }, { "epoch": 0.9430097951914514, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.23769760131836, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8565011620521545, "num_tokens": 282770491.0, "step": 7413 }, { "epoch": 0.943137005470042, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.257420539855957, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8578647375106812, "num_tokens": 282809824.0, "step": 7414 }, { "epoch": 0.9432642157486325, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.266276359558105, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8685617446899414, "num_tokens": 282846735.0, "step": 7415 }, { "epoch": 0.943391426027223, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.261510848999023, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8836065530776978, "num_tokens": 282882242.0, "step": 7416 }, { "epoch": 0.9435186363058136, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.282341957092285, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8518049716949463, "num_tokens": 282918774.0, "step": 7417 }, { "epoch": 0.943645846584404, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.2562894821167, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8564785718917847, "num_tokens": 282957766.0, "step": 7418 }, { "epoch": 0.9437730568629945, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.248659133911133, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8522730469703674, "num_tokens": 282995980.0, "step": 7419 }, { "epoch": 0.943900267141585, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.356161117553711, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8651261329650879, "num_tokens": 283033148.0, "step": 7420 }, { "epoch": 0.9440274774201756, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.191071510314941, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8694234490394592, "num_tokens": 283069781.0, "step": 7421 }, { "epoch": 0.9441546876987661, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.316819190979004, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8564785122871399, "num_tokens": 283107270.0, "step": 7422 }, { "epoch": 0.9442818979773566, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.294657707214355, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8770104646682739, "num_tokens": 283143571.0, "step": 7423 }, { "epoch": 0.944409108255947, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.29095458984375, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8681817054748535, "num_tokens": 283179995.0, "step": 7424 }, { "epoch": 0.9445363185345376, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.242030143737793, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8634934425354004, "num_tokens": 283218370.0, "step": 7425 }, { "epoch": 0.9446635288131281, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.325308799743652, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.857837438583374, "num_tokens": 283247879.0, "step": 7426 }, { "epoch": 0.9447907390917186, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.310798645019531, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8788085579872131, "num_tokens": 283285027.0, "step": 7427 }, { "epoch": 0.9449179493703092, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.205979347229004, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8672558069229126, "num_tokens": 283324934.0, "step": 7428 }, { "epoch": 0.9450451596488997, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.297858238220215, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8725579977035522, "num_tokens": 283357540.0, "step": 7429 }, { "epoch": 0.9451723699274901, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.265439987182617, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8618021011352539, "num_tokens": 283393470.0, "step": 7430 }, { "epoch": 0.9452995802060806, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.276789665222168, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8641519546508789, "num_tokens": 283429916.0, "step": 7431 }, { "epoch": 0.9454267904846712, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.271475791931152, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8538862466812134, "num_tokens": 283467997.0, "step": 7432 }, { "epoch": 0.9455540007632617, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.32220458984375, "learning_rate": 1e-06, "loss": 0.5224, "mean_token_accuracy": 0.830007016658783, "num_tokens": 283504741.0, "step": 7433 }, { "epoch": 0.9456812110418522, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.28367805480957, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8739814758300781, "num_tokens": 283534535.0, "step": 7434 }, { "epoch": 0.9458084213204427, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.306172370910645, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8753416538238525, "num_tokens": 283571600.0, "step": 7435 }, { "epoch": 0.9459356315990332, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.281399726867676, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8498299717903137, "num_tokens": 283605662.0, "step": 7436 }, { "epoch": 0.9460628418776237, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.232620239257812, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8641107082366943, "num_tokens": 283642868.0, "step": 7437 }, { "epoch": 0.9461900521562142, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.227540016174316, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8551602363586426, "num_tokens": 283675604.0, "step": 7438 }, { "epoch": 0.9463172624348047, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.230846405029297, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8843162059783936, "num_tokens": 283709459.0, "step": 7439 }, { "epoch": 0.9464444727133953, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.236482620239258, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8637705445289612, "num_tokens": 283748020.0, "step": 7440 }, { "epoch": 0.9465716829919858, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.291427612304688, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8592324256896973, "num_tokens": 283778514.0, "step": 7441 }, { "epoch": 0.9466988932705762, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.297465324401855, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8677990436553955, "num_tokens": 283819764.0, "step": 7442 }, { "epoch": 0.9468261035491667, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.274900436401367, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8500056266784668, "num_tokens": 283853897.0, "step": 7443 }, { "epoch": 0.9469533138277573, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.309331893920898, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8777130246162415, "num_tokens": 283891826.0, "step": 7444 }, { "epoch": 0.9470805241063478, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.278491020202637, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8623026609420776, "num_tokens": 283931690.0, "step": 7445 }, { "epoch": 0.9472077343849383, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.322078704833984, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8644582033157349, "num_tokens": 283971107.0, "step": 7446 }, { "epoch": 0.9473349446635289, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.267252922058105, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8714060187339783, "num_tokens": 284005286.0, "step": 7447 }, { "epoch": 0.9474621549421193, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.255636215209961, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8791201710700989, "num_tokens": 284044089.0, "step": 7448 }, { "epoch": 0.9475893652207098, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.324716567993164, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8597902655601501, "num_tokens": 284081172.0, "step": 7449 }, { "epoch": 0.9477165754993003, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.302541732788086, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8717485666275024, "num_tokens": 284113362.0, "step": 7450 }, { "epoch": 0.9478437857778909, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.289647102355957, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8736811876296997, "num_tokens": 284155272.0, "step": 7451 }, { "epoch": 0.9479709960564814, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.29974365234375, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8718376159667969, "num_tokens": 284188218.0, "step": 7452 }, { "epoch": 0.9480982063350719, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.290292739868164, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.867708146572113, "num_tokens": 284228838.0, "step": 7453 }, { "epoch": 0.9482254166136623, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.321496963500977, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8617949485778809, "num_tokens": 284265570.0, "step": 7454 }, { "epoch": 0.9483526268922529, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.243915557861328, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8542494177818298, "num_tokens": 284307062.0, "step": 7455 }, { "epoch": 0.9484798371708434, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.352102279663086, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8631905317306519, "num_tokens": 284340488.0, "step": 7456 }, { "epoch": 0.9486070474494339, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.283164024353027, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8635464906692505, "num_tokens": 284382399.0, "step": 7457 }, { "epoch": 0.9487342577280244, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.275274276733398, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8518512845039368, "num_tokens": 284416391.0, "step": 7458 }, { "epoch": 0.948861468006615, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.274412155151367, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8640652298927307, "num_tokens": 284457049.0, "step": 7459 }, { "epoch": 0.9489886782852054, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.217798233032227, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.862434446811676, "num_tokens": 284494426.0, "step": 7460 }, { "epoch": 0.9491158885637959, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.236746788024902, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8824884295463562, "num_tokens": 284533349.0, "step": 7461 }, { "epoch": 0.9492430988423864, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.257002830505371, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.872005045413971, "num_tokens": 284567777.0, "step": 7462 }, { "epoch": 0.949370309120977, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.29666519165039, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8585309982299805, "num_tokens": 284609120.0, "step": 7463 }, { "epoch": 0.9494975193995675, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.23682689666748, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8571856021881104, "num_tokens": 284649221.0, "step": 7464 }, { "epoch": 0.949624729678158, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.29272174835205, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8699064254760742, "num_tokens": 284682595.0, "step": 7465 }, { "epoch": 0.9497519399567486, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.267402648925781, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.86021488904953, "num_tokens": 284722903.0, "step": 7466 }, { "epoch": 0.949879150235339, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.315659523010254, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8691103458404541, "num_tokens": 284759166.0, "step": 7467 }, { "epoch": 0.9500063605139295, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.241106033325195, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8606851100921631, "num_tokens": 284803192.0, "step": 7468 }, { "epoch": 0.95013357079252, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.349343299865723, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8665168285369873, "num_tokens": 284837088.0, "step": 7469 }, { "epoch": 0.9502607810711106, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.249317169189453, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.8453602194786072, "num_tokens": 284879641.0, "step": 7470 }, { "epoch": 0.9503879913497011, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.281370162963867, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8537261486053467, "num_tokens": 284917519.0, "step": 7471 }, { "epoch": 0.9505152016282916, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.250842094421387, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8542108535766602, "num_tokens": 284950422.0, "step": 7472 }, { "epoch": 0.950642411906882, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.256052017211914, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.871537446975708, "num_tokens": 284986392.0, "step": 7473 }, { "epoch": 0.9507696221854726, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.25105094909668, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.865656852722168, "num_tokens": 285027684.0, "step": 7474 }, { "epoch": 0.9508968324640631, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.262556076049805, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8558931946754456, "num_tokens": 285059551.0, "step": 7475 }, { "epoch": 0.9510240427426536, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.259771347045898, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8652718663215637, "num_tokens": 285100197.0, "step": 7476 }, { "epoch": 0.9511512530212441, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.281652450561523, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8646194934844971, "num_tokens": 285130115.0, "step": 7477 }, { "epoch": 0.9512784632998347, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.303701400756836, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8649348020553589, "num_tokens": 285169069.0, "step": 7478 }, { "epoch": 0.9514056735784251, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.258951187133789, "learning_rate": 1e-06, "loss": 0.4992, "mean_token_accuracy": 0.8442558646202087, "num_tokens": 285209971.0, "step": 7479 }, { "epoch": 0.9515328838570156, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.246021270751953, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8594661951065063, "num_tokens": 285250024.0, "step": 7480 }, { "epoch": 0.9516600941356061, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.296195983886719, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8609703779220581, "num_tokens": 285284900.0, "step": 7481 }, { "epoch": 0.9517873044141967, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.281188011169434, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8676739931106567, "num_tokens": 285325657.0, "step": 7482 }, { "epoch": 0.9519145146927872, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.260748863220215, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8640040159225464, "num_tokens": 285356362.0, "step": 7483 }, { "epoch": 0.9520417249713777, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.26741886138916, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8634592294692993, "num_tokens": 285394321.0, "step": 7484 }, { "epoch": 0.9521689352499682, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.22995376586914, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8833926916122437, "num_tokens": 285435260.0, "step": 7485 }, { "epoch": 0.9522961455285587, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.304408073425293, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8720496892929077, "num_tokens": 285479449.0, "step": 7486 }, { "epoch": 0.9524233558071492, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.296744346618652, "learning_rate": 1e-06, "loss": 0.4965, "mean_token_accuracy": 0.8443912267684937, "num_tokens": 285520492.0, "step": 7487 }, { "epoch": 0.9525505660857397, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.2821044921875, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8606230616569519, "num_tokens": 285558410.0, "step": 7488 }, { "epoch": 0.9526777763643303, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.304193496704102, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8440243005752563, "num_tokens": 285598871.0, "step": 7489 }, { "epoch": 0.9528049866429208, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.296384811401367, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8575562238693237, "num_tokens": 285641374.0, "step": 7490 }, { "epoch": 0.9529321969215112, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.252975463867188, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8764042258262634, "num_tokens": 285675416.0, "step": 7491 }, { "epoch": 0.9530594072001017, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.346758842468262, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8708447813987732, "num_tokens": 285707491.0, "step": 7492 }, { "epoch": 0.9531866174786923, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.305002212524414, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8533672094345093, "num_tokens": 285747571.0, "step": 7493 }, { "epoch": 0.9533138277572828, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.260741233825684, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8857457637786865, "num_tokens": 285786362.0, "step": 7494 }, { "epoch": 0.9534410380358733, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.24421215057373, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8712630271911621, "num_tokens": 285829401.0, "step": 7495 }, { "epoch": 0.9535682483144639, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.25178337097168, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.861832857131958, "num_tokens": 285867853.0, "step": 7496 }, { "epoch": 0.9536954585930543, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.232880592346191, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8744845986366272, "num_tokens": 285908497.0, "step": 7497 }, { "epoch": 0.9538226688716448, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.2366943359375, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8512623310089111, "num_tokens": 285940528.0, "step": 7498 }, { "epoch": 0.9539498791502353, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.297693252563477, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8652259111404419, "num_tokens": 285980206.0, "step": 7499 }, { "epoch": 0.9540770894288259, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.217121124267578, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8578150272369385, "num_tokens": 286019409.0, "step": 7500 }, { "epoch": 0.9542042997074164, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.309679985046387, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8730603456497192, "num_tokens": 286050955.0, "step": 7501 }, { "epoch": 0.9543315099860069, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.238513946533203, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8690686821937561, "num_tokens": 286087567.0, "step": 7502 }, { "epoch": 0.9544587202645973, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.244446754455566, "learning_rate": 1e-06, "loss": 0.4881, "mean_token_accuracy": 0.8440326452255249, "num_tokens": 286126893.0, "step": 7503 }, { "epoch": 0.9545859305431879, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.252110481262207, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8569759130477905, "num_tokens": 286165818.0, "step": 7504 }, { "epoch": 0.9547131408217784, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.288578987121582, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8669595718383789, "num_tokens": 286208026.0, "step": 7505 }, { "epoch": 0.9548403511003689, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.257658004760742, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8637782335281372, "num_tokens": 286245468.0, "step": 7506 }, { "epoch": 0.9549675613789594, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.26390266418457, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8682972192764282, "num_tokens": 286283955.0, "step": 7507 }, { "epoch": 0.95509477165755, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.250452041625977, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.869726836681366, "num_tokens": 286326560.0, "step": 7508 }, { "epoch": 0.9552219819361404, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.324708938598633, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8646777272224426, "num_tokens": 286365791.0, "step": 7509 }, { "epoch": 0.9553491922147309, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.25027084350586, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8752907514572144, "num_tokens": 286404132.0, "step": 7510 }, { "epoch": 0.9554764024933214, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.272500038146973, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8540374636650085, "num_tokens": 286444828.0, "step": 7511 }, { "epoch": 0.955603612771912, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.314050674438477, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8690161108970642, "num_tokens": 286481725.0, "step": 7512 }, { "epoch": 0.9557308230505025, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.219999313354492, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.864754319190979, "num_tokens": 286520354.0, "step": 7513 }, { "epoch": 0.955858033329093, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.252511978149414, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8617175817489624, "num_tokens": 286564314.0, "step": 7514 }, { "epoch": 0.9559852436076836, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.2652587890625, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8635365962982178, "num_tokens": 286606955.0, "step": 7515 }, { "epoch": 0.956112453886274, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.285425186157227, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8708276152610779, "num_tokens": 286640091.0, "step": 7516 }, { "epoch": 0.9562396641648645, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.346050262451172, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8636856079101562, "num_tokens": 286679090.0, "step": 7517 }, { "epoch": 0.956366874443455, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.225544929504395, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8688187599182129, "num_tokens": 286720260.0, "step": 7518 }, { "epoch": 0.9564940847220456, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.274731636047363, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8701865077018738, "num_tokens": 286763992.0, "step": 7519 }, { "epoch": 0.9566212950006361, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.24726676940918, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8453352451324463, "num_tokens": 286807068.0, "step": 7520 }, { "epoch": 0.9567485052792266, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24863338470459, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8706343173980713, "num_tokens": 286853020.0, "step": 7521 }, { "epoch": 0.956875715557817, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.289836883544922, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8540120124816895, "num_tokens": 286896098.0, "step": 7522 }, { "epoch": 0.9570029258364076, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.284624099731445, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8487482070922852, "num_tokens": 286938873.0, "step": 7523 }, { "epoch": 0.9571301361149981, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.268180847167969, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8686400651931763, "num_tokens": 286972941.0, "step": 7524 }, { "epoch": 0.9572573463935886, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.299808502197266, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8537582755088806, "num_tokens": 287010818.0, "step": 7525 }, { "epoch": 0.9573845566721791, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.244566917419434, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8684262037277222, "num_tokens": 287053841.0, "step": 7526 }, { "epoch": 0.9575117669507697, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.270828247070312, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8695955276489258, "num_tokens": 287090900.0, "step": 7527 }, { "epoch": 0.9576389772293601, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.24075984954834, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8628999590873718, "num_tokens": 287131203.0, "step": 7528 }, { "epoch": 0.9577661875079506, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.31613540649414, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8757112622261047, "num_tokens": 287165949.0, "step": 7529 }, { "epoch": 0.9578933977865411, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.279507637023926, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8645904064178467, "num_tokens": 287203376.0, "step": 7530 }, { "epoch": 0.9580206080651317, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.215109825134277, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8552284240722656, "num_tokens": 287240391.0, "step": 7531 }, { "epoch": 0.9581478183437222, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.24354362487793, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8613743782043457, "num_tokens": 287278485.0, "step": 7532 }, { "epoch": 0.9582750286223127, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.356773376464844, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8589897155761719, "num_tokens": 287313626.0, "step": 7533 }, { "epoch": 0.9584022389009031, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.242012023925781, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8796570897102356, "num_tokens": 287344532.0, "step": 7534 }, { "epoch": 0.9585294491794937, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.325258255004883, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.84559565782547, "num_tokens": 287383069.0, "step": 7535 }, { "epoch": 0.9586566594580842, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.279499053955078, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8638801574707031, "num_tokens": 287418554.0, "step": 7536 }, { "epoch": 0.9587838697366747, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.299351692199707, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.86298668384552, "num_tokens": 287452449.0, "step": 7537 }, { "epoch": 0.9589110800152653, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.311178207397461, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8500096797943115, "num_tokens": 287492247.0, "step": 7538 }, { "epoch": 0.9590382902938558, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.254817008972168, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8701111674308777, "num_tokens": 287528651.0, "step": 7539 }, { "epoch": 0.9591655005724462, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.278697967529297, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8545129299163818, "num_tokens": 287565837.0, "step": 7540 }, { "epoch": 0.9592927108510367, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.304573059082031, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8633226156234741, "num_tokens": 287599469.0, "step": 7541 }, { "epoch": 0.9594199211296273, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.251726150512695, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8618288040161133, "num_tokens": 287637104.0, "step": 7542 }, { "epoch": 0.9595471314082178, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.249361038208008, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8644004464149475, "num_tokens": 287679134.0, "step": 7543 }, { "epoch": 0.9596743416868083, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.221392631530762, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8718196153640747, "num_tokens": 287720340.0, "step": 7544 }, { "epoch": 0.9598015519653988, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.22149658203125, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8684781789779663, "num_tokens": 287760931.0, "step": 7545 }, { "epoch": 0.9599287622439893, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.273750305175781, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8508843183517456, "num_tokens": 287801723.0, "step": 7546 }, { "epoch": 0.9600559725225798, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.275659561157227, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8627979755401611, "num_tokens": 287837010.0, "step": 7547 }, { "epoch": 0.9601831828011703, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.265311241149902, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8662532567977905, "num_tokens": 287874381.0, "step": 7548 }, { "epoch": 0.9603103930797608, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.311941146850586, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.856788694858551, "num_tokens": 287916673.0, "step": 7549 }, { "epoch": 0.9604376033583514, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.303985595703125, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8592992424964905, "num_tokens": 287956764.0, "step": 7550 }, { "epoch": 0.9605648136369419, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.262444496154785, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.87189120054245, "num_tokens": 287996373.0, "step": 7551 }, { "epoch": 0.9606920239155323, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.33236026763916, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8583843111991882, "num_tokens": 288033671.0, "step": 7552 }, { "epoch": 0.9608192341941229, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.20274543762207, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8728883266448975, "num_tokens": 288068254.0, "step": 7553 }, { "epoch": 0.9609464444727134, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.309581756591797, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8497270345687866, "num_tokens": 288102833.0, "step": 7554 }, { "epoch": 0.9610736547513039, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.320845603942871, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8769462704658508, "num_tokens": 288135927.0, "step": 7555 }, { "epoch": 0.9612008650298944, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.233349800109863, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8832048177719116, "num_tokens": 288173842.0, "step": 7556 }, { "epoch": 0.961328075308485, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.308492660522461, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8696779012680054, "num_tokens": 288215007.0, "step": 7557 }, { "epoch": 0.9614552855870754, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.315977096557617, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8534383177757263, "num_tokens": 288251598.0, "step": 7558 }, { "epoch": 0.9615824958656659, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.226301193237305, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8650219440460205, "num_tokens": 288292811.0, "step": 7559 }, { "epoch": 0.9617097061442564, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.264501571655273, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.869264543056488, "num_tokens": 288337790.0, "step": 7560 }, { "epoch": 0.961836916422847, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.335197448730469, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8621050119400024, "num_tokens": 288373347.0, "step": 7561 }, { "epoch": 0.9619641267014375, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.29018783569336, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8844494819641113, "num_tokens": 288405650.0, "step": 7562 }, { "epoch": 0.962091336980028, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.26267147064209, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8706218600273132, "num_tokens": 288443849.0, "step": 7563 }, { "epoch": 0.9622185472586186, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.292709350585938, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8430991172790527, "num_tokens": 288487557.0, "step": 7564 }, { "epoch": 0.962345757537209, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.277857780456543, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8551113605499268, "num_tokens": 288529932.0, "step": 7565 }, { "epoch": 0.9624729678157995, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.27613639831543, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.859827995300293, "num_tokens": 288568276.0, "step": 7566 }, { "epoch": 0.96260017809439, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.319022178649902, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8663439750671387, "num_tokens": 288602387.0, "step": 7567 }, { "epoch": 0.9627273883729806, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.311201095581055, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8719711303710938, "num_tokens": 288638157.0, "step": 7568 }, { "epoch": 0.9628545986515711, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.310263633728027, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8611443638801575, "num_tokens": 288675054.0, "step": 7569 }, { "epoch": 0.9629818089301616, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.253286361694336, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8695212006568909, "num_tokens": 288714222.0, "step": 7570 }, { "epoch": 0.963109019208752, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.297673225402832, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8538064360618591, "num_tokens": 288752772.0, "step": 7571 }, { "epoch": 0.9632362294873426, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.266698837280273, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8479714393615723, "num_tokens": 288792067.0, "step": 7572 }, { "epoch": 0.9633634397659331, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.386883735656738, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8761334419250488, "num_tokens": 288826172.0, "step": 7573 }, { "epoch": 0.9634906500445236, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.276618003845215, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8738564252853394, "num_tokens": 288863774.0, "step": 7574 }, { "epoch": 0.9636178603231141, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.305253028869629, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8582838177680969, "num_tokens": 288901892.0, "step": 7575 }, { "epoch": 0.9637450706017047, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.30477237701416, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8543028831481934, "num_tokens": 288941470.0, "step": 7576 }, { "epoch": 0.9638722808802951, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.287999153137207, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8509128093719482, "num_tokens": 288988616.0, "step": 7577 }, { "epoch": 0.9639994911588856, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.322530746459961, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8800762891769409, "num_tokens": 289024589.0, "step": 7578 }, { "epoch": 0.9641267014374761, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.340285301208496, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8668286204338074, "num_tokens": 289064334.0, "step": 7579 }, { "epoch": 0.9642539117160667, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.31502914428711, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.853275716304779, "num_tokens": 289099385.0, "step": 7580 }, { "epoch": 0.9643811219946572, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.29878044128418, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8676127195358276, "num_tokens": 289139143.0, "step": 7581 }, { "epoch": 0.9645083322732477, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.374454498291016, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8582056760787964, "num_tokens": 289179303.0, "step": 7582 }, { "epoch": 0.9646355425518381, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.365979194641113, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8569238185882568, "num_tokens": 289214218.0, "step": 7583 }, { "epoch": 0.9647627528304287, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.381608963012695, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8599897623062134, "num_tokens": 289240991.0, "step": 7584 }, { "epoch": 0.9648899631090192, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.291595458984375, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8575080633163452, "num_tokens": 289273374.0, "step": 7585 }, { "epoch": 0.9650171733876097, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.28537368774414, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8502631187438965, "num_tokens": 289314904.0, "step": 7586 }, { "epoch": 0.9651443836662003, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.373821258544922, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8571355938911438, "num_tokens": 289358110.0, "step": 7587 }, { "epoch": 0.9652715939447908, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.298892974853516, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8545029163360596, "num_tokens": 289391581.0, "step": 7588 }, { "epoch": 0.9653988042233812, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.349129676818848, "learning_rate": 1e-06, "loss": 0.4979, "mean_token_accuracy": 0.8393512964248657, "num_tokens": 289432280.0, "step": 7589 }, { "epoch": 0.9655260145019717, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.337857246398926, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8480038642883301, "num_tokens": 289473222.0, "step": 7590 }, { "epoch": 0.9656532247805623, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.374897956848145, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8597327470779419, "num_tokens": 289512898.0, "step": 7591 }, { "epoch": 0.9657804350591528, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.295366287231445, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8694077730178833, "num_tokens": 289554394.0, "step": 7592 }, { "epoch": 0.9659076453377433, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.373906135559082, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8600069880485535, "num_tokens": 289594871.0, "step": 7593 }, { "epoch": 0.9660348556163338, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.310612678527832, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8526986837387085, "num_tokens": 289629165.0, "step": 7594 }, { "epoch": 0.9661620658949243, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.281248092651367, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8624184131622314, "num_tokens": 289670131.0, "step": 7595 }, { "epoch": 0.9662892761735148, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.387299537658691, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8670268058776855, "num_tokens": 289705238.0, "step": 7596 }, { "epoch": 0.9664164864521053, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.386507987976074, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.86661696434021, "num_tokens": 289740146.0, "step": 7597 }, { "epoch": 0.9665436967306958, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.296504974365234, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8683139085769653, "num_tokens": 289781145.0, "step": 7598 }, { "epoch": 0.9666709070092864, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.374394416809082, "learning_rate": 1e-06, "loss": 0.4777, "mean_token_accuracy": 0.8465242385864258, "num_tokens": 289816347.0, "step": 7599 }, { "epoch": 0.9667981172878769, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.26329231262207, "learning_rate": 1e-06, "loss": 0.4847, "mean_token_accuracy": 0.8493931293487549, "num_tokens": 289852974.0, "step": 7600 }, { "epoch": 0.9669253275664673, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.307909965515137, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8516649603843689, "num_tokens": 289894553.0, "step": 7601 }, { "epoch": 0.9670525378450578, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.357307434082031, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.845771312713623, "num_tokens": 289930641.0, "step": 7602 }, { "epoch": 0.9671797481236484, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.347208976745605, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8577866554260254, "num_tokens": 289972690.0, "step": 7603 }, { "epoch": 0.9673069584022389, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.351798057556152, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.857874870300293, "num_tokens": 290006518.0, "step": 7604 }, { "epoch": 0.9674341686808294, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.343021392822266, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8505435585975647, "num_tokens": 290045206.0, "step": 7605 }, { "epoch": 0.96756137895942, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.299479484558105, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8683458566665649, "num_tokens": 290086976.0, "step": 7606 }, { "epoch": 0.9676885892380104, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.337688446044922, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8762336373329163, "num_tokens": 290126894.0, "step": 7607 }, { "epoch": 0.9678157995166009, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.354153633117676, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.876690685749054, "num_tokens": 290161684.0, "step": 7608 }, { "epoch": 0.9679430097951914, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.329386711120605, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8656137585639954, "num_tokens": 290203407.0, "step": 7609 }, { "epoch": 0.968070220073782, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.336752891540527, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8695561289787292, "num_tokens": 290238645.0, "step": 7610 }, { "epoch": 0.9681974303523725, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.280935287475586, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8625510334968567, "num_tokens": 290280279.0, "step": 7611 }, { "epoch": 0.968324640630963, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.358678817749023, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.862296998500824, "num_tokens": 290317739.0, "step": 7612 }, { "epoch": 0.9684518509095535, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.370650291442871, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8606812357902527, "num_tokens": 290360384.0, "step": 7613 }, { "epoch": 0.968579061188144, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.3777437210083, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8710282444953918, "num_tokens": 290398706.0, "step": 7614 }, { "epoch": 0.9687062714667345, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.294801712036133, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8772480487823486, "num_tokens": 290439524.0, "step": 7615 }, { "epoch": 0.968833481745325, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.317610740661621, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8584478497505188, "num_tokens": 290472125.0, "step": 7616 }, { "epoch": 0.9689606920239155, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.38093090057373, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8479104042053223, "num_tokens": 290514348.0, "step": 7617 }, { "epoch": 0.9690879023025061, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.256235122680664, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8582543134689331, "num_tokens": 290551699.0, "step": 7618 }, { "epoch": 0.9692151125810966, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.37365436553955, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8655514717102051, "num_tokens": 290588644.0, "step": 7619 }, { "epoch": 0.969342322859687, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.303084373474121, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8657835125923157, "num_tokens": 290629546.0, "step": 7620 }, { "epoch": 0.9694695331382776, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.32621955871582, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8720188140869141, "num_tokens": 290665527.0, "step": 7621 }, { "epoch": 0.9695967434168681, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.310608863830566, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8551933169364929, "num_tokens": 290706667.0, "step": 7622 }, { "epoch": 0.9697239536954586, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.320014953613281, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8585470914840698, "num_tokens": 290745698.0, "step": 7623 }, { "epoch": 0.9698511639740491, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.396289825439453, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8688740730285645, "num_tokens": 290777983.0, "step": 7624 }, { "epoch": 0.9699783742526397, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.2542085647583, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8566468358039856, "num_tokens": 290814184.0, "step": 7625 }, { "epoch": 0.9701055845312301, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.362749099731445, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8542119860649109, "num_tokens": 290854054.0, "step": 7626 }, { "epoch": 0.9702327948098206, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.307100296020508, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8623926639556885, "num_tokens": 290891325.0, "step": 7627 }, { "epoch": 0.9703600050884111, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.292142868041992, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8621131777763367, "num_tokens": 290930101.0, "step": 7628 }, { "epoch": 0.9704872153670017, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.36325740814209, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8631654977798462, "num_tokens": 290964691.0, "step": 7629 }, { "epoch": 0.9706144256455922, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.346054077148438, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8590323328971863, "num_tokens": 291008140.0, "step": 7630 }, { "epoch": 0.9707416359241827, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.313741683959961, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8574500679969788, "num_tokens": 291041484.0, "step": 7631 }, { "epoch": 0.9708688462027731, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.368061065673828, "learning_rate": 1e-06, "loss": 0.491, "mean_token_accuracy": 0.8438581228256226, "num_tokens": 291082337.0, "step": 7632 }, { "epoch": 0.9709960564813637, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.290361404418945, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8627873659133911, "num_tokens": 291115921.0, "step": 7633 }, { "epoch": 0.9711232667599542, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.305790901184082, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8711936473846436, "num_tokens": 291154367.0, "step": 7634 }, { "epoch": 0.9712504770385447, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.305708885192871, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8635216951370239, "num_tokens": 291198919.0, "step": 7635 }, { "epoch": 0.9713776873171353, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.386412620544434, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8516620397567749, "num_tokens": 291237062.0, "step": 7636 }, { "epoch": 0.9715048975957258, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.320332527160645, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8744610548019409, "num_tokens": 291275676.0, "step": 7637 }, { "epoch": 0.9716321078743162, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.369383811950684, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.855635404586792, "num_tokens": 291316785.0, "step": 7638 }, { "epoch": 0.9717593181529067, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.306790351867676, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8608843684196472, "num_tokens": 291360448.0, "step": 7639 }, { "epoch": 0.9718865284314973, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.319478988647461, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8596129417419434, "num_tokens": 291401650.0, "step": 7640 }, { "epoch": 0.9720137387100878, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.30256462097168, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8636784553527832, "num_tokens": 291439736.0, "step": 7641 }, { "epoch": 0.9721409489886783, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.296456336975098, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8691638112068176, "num_tokens": 291479077.0, "step": 7642 }, { "epoch": 0.9722681592672688, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.378808975219727, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8734241724014282, "num_tokens": 291512470.0, "step": 7643 }, { "epoch": 0.9723953695458593, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.293777465820312, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8734925389289856, "num_tokens": 291552797.0, "step": 7644 }, { "epoch": 0.9725225798244498, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.385826110839844, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.866521954536438, "num_tokens": 291589885.0, "step": 7645 }, { "epoch": 0.9726497901030403, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.298971176147461, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.862744927406311, "num_tokens": 291633005.0, "step": 7646 }, { "epoch": 0.9727770003816308, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.318150520324707, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.860491156578064, "num_tokens": 291666083.0, "step": 7647 }, { "epoch": 0.9729042106602214, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.348001480102539, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8692013025283813, "num_tokens": 291706421.0, "step": 7648 }, { "epoch": 0.9730314209388119, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.294610977172852, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8674938678741455, "num_tokens": 291744641.0, "step": 7649 }, { "epoch": 0.9731586312174023, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.303363800048828, "learning_rate": 1e-06, "loss": 0.3425, "mean_token_accuracy": 0.8893760442733765, "num_tokens": 291780586.0, "step": 7650 }, { "epoch": 0.9732858414959928, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.312272071838379, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8588260412216187, "num_tokens": 291820482.0, "step": 7651 }, { "epoch": 0.9734130517745834, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.33403205871582, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8723496198654175, "num_tokens": 291858568.0, "step": 7652 }, { "epoch": 0.9735402620531739, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.371742248535156, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8547363877296448, "num_tokens": 291889383.0, "step": 7653 }, { "epoch": 0.9736674723317644, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.36616039276123, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8719396591186523, "num_tokens": 291927541.0, "step": 7654 }, { "epoch": 0.973794682610355, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.362528800964355, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8477123975753784, "num_tokens": 291969074.0, "step": 7655 }, { "epoch": 0.9739218928889454, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.287077903747559, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8618168830871582, "num_tokens": 292008311.0, "step": 7656 }, { "epoch": 0.9740491031675359, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.417401313781738, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8544512987136841, "num_tokens": 292051810.0, "step": 7657 }, { "epoch": 0.9741763134461264, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.28159236907959, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8706513047218323, "num_tokens": 292090411.0, "step": 7658 }, { "epoch": 0.974303523724717, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.36274528503418, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.856070876121521, "num_tokens": 292124411.0, "step": 7659 }, { "epoch": 0.9744307340033075, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.299626350402832, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8635094165802002, "num_tokens": 292166428.0, "step": 7660 }, { "epoch": 0.974557944281898, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.337817192077637, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8654314279556274, "num_tokens": 292202687.0, "step": 7661 }, { "epoch": 0.9746851545604885, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.305683135986328, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8733923435211182, "num_tokens": 292246364.0, "step": 7662 }, { "epoch": 0.974812364839079, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.308685302734375, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8640899658203125, "num_tokens": 292285789.0, "step": 7663 }, { "epoch": 0.9749395751176695, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.371570587158203, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8582369089126587, "num_tokens": 292320755.0, "step": 7664 }, { "epoch": 0.97506678539626, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.247255325317383, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8726189136505127, "num_tokens": 292362818.0, "step": 7665 }, { "epoch": 0.9751939956748505, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.334197044372559, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8560713529586792, "num_tokens": 292404016.0, "step": 7666 }, { "epoch": 0.9753212059534411, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.337114334106445, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8501369953155518, "num_tokens": 292442117.0, "step": 7667 }, { "epoch": 0.9754484162320316, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.319113731384277, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8714547157287598, "num_tokens": 292472523.0, "step": 7668 }, { "epoch": 0.975575626510622, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.356831550598145, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.867792010307312, "num_tokens": 292508571.0, "step": 7669 }, { "epoch": 0.9757028367892125, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.26527214050293, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8598792552947998, "num_tokens": 292545358.0, "step": 7670 }, { "epoch": 0.9758300470678031, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.383462905883789, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8487660884857178, "num_tokens": 292583962.0, "step": 7671 }, { "epoch": 0.9759572573463936, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.317376136779785, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8625808358192444, "num_tokens": 292616186.0, "step": 7672 }, { "epoch": 0.9760844676249841, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.323996543884277, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8594602346420288, "num_tokens": 292657378.0, "step": 7673 }, { "epoch": 0.9762116779035747, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.316462516784668, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8769670724868774, "num_tokens": 292692041.0, "step": 7674 }, { "epoch": 0.9763388881821651, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.331625938415527, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8611768484115601, "num_tokens": 292725062.0, "step": 7675 }, { "epoch": 0.9764660984607556, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.2696533203125, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.871116578578949, "num_tokens": 292766522.0, "step": 7676 }, { "epoch": 0.9765933087393461, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.346442222595215, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8631739616394043, "num_tokens": 292811371.0, "step": 7677 }, { "epoch": 0.9767205190179367, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.262045860290527, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8639386296272278, "num_tokens": 292850237.0, "step": 7678 }, { "epoch": 0.9768477292965272, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.327136039733887, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8668962717056274, "num_tokens": 292887291.0, "step": 7679 }, { "epoch": 0.9769749395751177, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.296439170837402, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8758935928344727, "num_tokens": 292928396.0, "step": 7680 }, { "epoch": 0.9771021498537081, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.339865684509277, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8671391010284424, "num_tokens": 292962748.0, "step": 7681 }, { "epoch": 0.9772293601322987, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.322066307067871, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8751622438430786, "num_tokens": 293002330.0, "step": 7682 }, { "epoch": 0.9773565704108892, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.327779769897461, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8633787035942078, "num_tokens": 293038288.0, "step": 7683 }, { "epoch": 0.9774837806894797, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.274799346923828, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8666887879371643, "num_tokens": 293078121.0, "step": 7684 }, { "epoch": 0.9776109909680702, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.334545135498047, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8605554103851318, "num_tokens": 293116846.0, "step": 7685 }, { "epoch": 0.9777382012466608, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.351872444152832, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8588281869888306, "num_tokens": 293162237.0, "step": 7686 }, { "epoch": 0.9778654115252512, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.313065528869629, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8603506088256836, "num_tokens": 293196631.0, "step": 7687 }, { "epoch": 0.9779926218038417, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.376760482788086, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8737679719924927, "num_tokens": 293238883.0, "step": 7688 }, { "epoch": 0.9781198320824323, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.370786666870117, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8534114956855774, "num_tokens": 293270802.0, "step": 7689 }, { "epoch": 0.9782470423610228, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.368677139282227, "learning_rate": 1e-06, "loss": 0.4928, "mean_token_accuracy": 0.8496609330177307, "num_tokens": 293314910.0, "step": 7690 }, { "epoch": 0.9783742526396133, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.320024490356445, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8760160207748413, "num_tokens": 293353083.0, "step": 7691 }, { "epoch": 0.9785014629182038, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.35362434387207, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8651214838027954, "num_tokens": 293389473.0, "step": 7692 }, { "epoch": 0.9786286731967943, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.366164207458496, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8762339949607849, "num_tokens": 293426893.0, "step": 7693 }, { "epoch": 0.9787558834753848, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.322064399719238, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8691606521606445, "num_tokens": 293468116.0, "step": 7694 }, { "epoch": 0.9788830937539753, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.369515419006348, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8666456341743469, "num_tokens": 293503050.0, "step": 7695 }, { "epoch": 0.9790103040325658, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.383545875549316, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.853401780128479, "num_tokens": 293541673.0, "step": 7696 }, { "epoch": 0.9791375143111564, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.328773498535156, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8612099885940552, "num_tokens": 293578472.0, "step": 7697 }, { "epoch": 0.9792647245897469, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.381784439086914, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8665649890899658, "num_tokens": 293611204.0, "step": 7698 }, { "epoch": 0.9793919348683373, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.350605010986328, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.854002833366394, "num_tokens": 293651352.0, "step": 7699 }, { "epoch": 0.9795191451469278, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.313389778137207, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8737590909004211, "num_tokens": 293687396.0, "step": 7700 }, { "epoch": 0.9796463554255184, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.385396957397461, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8593956232070923, "num_tokens": 293725083.0, "step": 7701 }, { "epoch": 0.9797735657041089, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.3136568069458, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8643518686294556, "num_tokens": 293766136.0, "step": 7702 }, { "epoch": 0.9799007759826994, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.40573787689209, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8471829891204834, "num_tokens": 293805810.0, "step": 7703 }, { "epoch": 0.98002798626129, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.326809883117676, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8556463718414307, "num_tokens": 293839041.0, "step": 7704 }, { "epoch": 0.9801551965398804, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.38276195526123, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8789168000221252, "num_tokens": 293876541.0, "step": 7705 }, { "epoch": 0.9802824068184709, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.303771018981934, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8777209520339966, "num_tokens": 293918141.0, "step": 7706 }, { "epoch": 0.9804096170970614, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.477251052856445, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8524975180625916, "num_tokens": 293958078.0, "step": 7707 }, { "epoch": 0.980536827375652, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.389512062072754, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8734495639801025, "num_tokens": 293990347.0, "step": 7708 }, { "epoch": 0.9806640376542425, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.426973342895508, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.868722677230835, "num_tokens": 294032136.0, "step": 7709 }, { "epoch": 0.980791247932833, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.403918266296387, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8512505292892456, "num_tokens": 294065136.0, "step": 7710 }, { "epoch": 0.9809184582114235, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.409777641296387, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8781383633613586, "num_tokens": 294104539.0, "step": 7711 }, { "epoch": 0.981045668490014, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.289368629455566, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.861099123954773, "num_tokens": 294148962.0, "step": 7712 }, { "epoch": 0.9811728787686045, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.35159969329834, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8677363395690918, "num_tokens": 294181866.0, "step": 7713 }, { "epoch": 0.981300089047195, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.37846851348877, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8724966645240784, "num_tokens": 294223558.0, "step": 7714 }, { "epoch": 0.9814272993257855, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.258831977844238, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8682923316955566, "num_tokens": 294260487.0, "step": 7715 }, { "epoch": 0.9815545096043761, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.396905899047852, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8654492497444153, "num_tokens": 294302642.0, "step": 7716 }, { "epoch": 0.9816817198829666, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.286336898803711, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8553575277328491, "num_tokens": 294338376.0, "step": 7717 }, { "epoch": 0.981808930161557, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.373982429504395, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8587636947631836, "num_tokens": 294372089.0, "step": 7718 }, { "epoch": 0.9819361404401475, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.32772159576416, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8693393468856812, "num_tokens": 294403801.0, "step": 7719 }, { "epoch": 0.9820633507187381, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.359134674072266, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8668295741081238, "num_tokens": 294440301.0, "step": 7720 }, { "epoch": 0.9821905609973286, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.369667053222656, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8637077808380127, "num_tokens": 294472967.0, "step": 7721 }, { "epoch": 0.9823177712759191, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.34677791595459, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8550474643707275, "num_tokens": 294518248.0, "step": 7722 }, { "epoch": 0.9824449815545097, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.37354850769043, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8625900745391846, "num_tokens": 294553899.0, "step": 7723 }, { "epoch": 0.9825721918331001, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.368101119995117, "learning_rate": 1e-06, "loss": 0.4804, "mean_token_accuracy": 0.8418371677398682, "num_tokens": 294591382.0, "step": 7724 }, { "epoch": 0.9826994021116906, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.375927925109863, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8645024299621582, "num_tokens": 294629606.0, "step": 7725 }, { "epoch": 0.9828266123902811, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.365570068359375, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8660807609558105, "num_tokens": 294669417.0, "step": 7726 }, { "epoch": 0.9829538226688717, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.41689682006836, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8732575178146362, "num_tokens": 294709876.0, "step": 7727 }, { "epoch": 0.9830810329474622, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.31169605255127, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.855307400226593, "num_tokens": 294750565.0, "step": 7728 }, { "epoch": 0.9832082432260527, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.36713695526123, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8642244338989258, "num_tokens": 294790586.0, "step": 7729 }, { "epoch": 0.9833354535046431, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.398786544799805, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.862603485584259, "num_tokens": 294830176.0, "step": 7730 }, { "epoch": 0.9834626637832337, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.293717384338379, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8522230982780457, "num_tokens": 294874199.0, "step": 7731 }, { "epoch": 0.9835898740618242, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.411894798278809, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8707795143127441, "num_tokens": 294914219.0, "step": 7732 }, { "epoch": 0.9837170843404147, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.366098403930664, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8415358066558838, "num_tokens": 294950726.0, "step": 7733 }, { "epoch": 0.9838442946190052, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.373310089111328, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8640161752700806, "num_tokens": 294993815.0, "step": 7734 }, { "epoch": 0.9839715048975958, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.337091445922852, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8723196983337402, "num_tokens": 295029275.0, "step": 7735 }, { "epoch": 0.9840987151761862, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.360795021057129, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8628933429718018, "num_tokens": 295066166.0, "step": 7736 }, { "epoch": 0.9842259254547767, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.404705047607422, "learning_rate": 1e-06, "loss": 0.5029, "mean_token_accuracy": 0.841620922088623, "num_tokens": 295100436.0, "step": 7737 }, { "epoch": 0.9843531357333672, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.305685043334961, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8652271628379822, "num_tokens": 295141756.0, "step": 7738 }, { "epoch": 0.9844803460119578, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.370305061340332, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8807770013809204, "num_tokens": 295178894.0, "step": 7739 }, { "epoch": 0.9846075562905483, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.328385353088379, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8532522916793823, "num_tokens": 295219133.0, "step": 7740 }, { "epoch": 0.9847347665691388, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.397989273071289, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8642731308937073, "num_tokens": 295260322.0, "step": 7741 }, { "epoch": 0.9848619768477292, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.416882514953613, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8585262298583984, "num_tokens": 295299675.0, "step": 7742 }, { "epoch": 0.9849891871263198, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.356268882751465, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8695805072784424, "num_tokens": 295336281.0, "step": 7743 }, { "epoch": 0.9851163974049103, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.348445892333984, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8610380291938782, "num_tokens": 295376932.0, "step": 7744 }, { "epoch": 0.9852436076835008, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.388541221618652, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8600375652313232, "num_tokens": 295421904.0, "step": 7745 }, { "epoch": 0.9853708179620914, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.36032772064209, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.866234540939331, "num_tokens": 295456688.0, "step": 7746 }, { "epoch": 0.9854980282406819, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.354708671569824, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8807595372200012, "num_tokens": 295500022.0, "step": 7747 }, { "epoch": 0.9856252385192723, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.379385948181152, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.883643627166748, "num_tokens": 295538106.0, "step": 7748 }, { "epoch": 0.9857524487978628, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.405241966247559, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8595381379127502, "num_tokens": 295572465.0, "step": 7749 }, { "epoch": 0.9858796590764534, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.34109115600586, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8696874380111694, "num_tokens": 295614577.0, "step": 7750 }, { "epoch": 0.9860068693550439, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.37918472290039, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8681453466415405, "num_tokens": 295654780.0, "step": 7751 }, { "epoch": 0.9861340796336344, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.340888977050781, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8627142906188965, "num_tokens": 295695880.0, "step": 7752 }, { "epoch": 0.986261289912225, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.438142776489258, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8633718490600586, "num_tokens": 295738176.0, "step": 7753 }, { "epoch": 0.9863885001908154, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.386136054992676, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.853752613067627, "num_tokens": 295777562.0, "step": 7754 }, { "epoch": 0.9865157104694059, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.330520629882812, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8806425929069519, "num_tokens": 295815013.0, "step": 7755 }, { "epoch": 0.9866429207479964, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.403643608093262, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8522013425827026, "num_tokens": 295857751.0, "step": 7756 }, { "epoch": 0.986770131026587, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.362628936767578, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8687872886657715, "num_tokens": 295896033.0, "step": 7757 }, { "epoch": 0.9868973413051775, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.405083656311035, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8661826252937317, "num_tokens": 295936838.0, "step": 7758 }, { "epoch": 0.987024551583768, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.325665473937988, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8698900938034058, "num_tokens": 295972307.0, "step": 7759 }, { "epoch": 0.9871517618623584, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.335277557373047, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.86161869764328, "num_tokens": 296009966.0, "step": 7760 }, { "epoch": 0.987278972140949, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.385334014892578, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8613100051879883, "num_tokens": 296047874.0, "step": 7761 }, { "epoch": 0.9874061824195395, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.323548316955566, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8713178634643555, "num_tokens": 296090591.0, "step": 7762 }, { "epoch": 0.98753339269813, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.355289459228516, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8588166236877441, "num_tokens": 296133024.0, "step": 7763 }, { "epoch": 0.9876606029767205, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.341086387634277, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8530975580215454, "num_tokens": 296177347.0, "step": 7764 }, { "epoch": 0.9877878132553111, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.38211441040039, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8541645407676697, "num_tokens": 296215532.0, "step": 7765 }, { "epoch": 0.9879150235339016, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.422005653381348, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8531795740127563, "num_tokens": 296249494.0, "step": 7766 }, { "epoch": 0.988042233812492, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.389654159545898, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8549657464027405, "num_tokens": 296289370.0, "step": 7767 }, { "epoch": 0.9881694440910825, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.364288330078125, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8644660711288452, "num_tokens": 296329376.0, "step": 7768 }, { "epoch": 0.9882966543696731, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.380697250366211, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8712272644042969, "num_tokens": 296368875.0, "step": 7769 }, { "epoch": 0.9884238646482636, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.389971733093262, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8626399040222168, "num_tokens": 296404766.0, "step": 7770 }, { "epoch": 0.9885510749268541, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.341377258300781, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8633782267570496, "num_tokens": 296441705.0, "step": 7771 }, { "epoch": 0.9886782852054447, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.431132316589355, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8709098100662231, "num_tokens": 296481075.0, "step": 7772 }, { "epoch": 0.9888054954840351, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.421577453613281, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8600611686706543, "num_tokens": 296511404.0, "step": 7773 }, { "epoch": 0.9889327057626256, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.398539543151855, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8542156219482422, "num_tokens": 296546080.0, "step": 7774 }, { "epoch": 0.9890599160412161, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.319098472595215, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8660104274749756, "num_tokens": 296586119.0, "step": 7775 }, { "epoch": 0.9891871263198067, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.354401588439941, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8626599311828613, "num_tokens": 296629602.0, "step": 7776 }, { "epoch": 0.9893143365983972, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.385709762573242, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8573758006095886, "num_tokens": 296666709.0, "step": 7777 }, { "epoch": 0.9894415468769877, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.322147369384766, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8634947538375854, "num_tokens": 296702611.0, "step": 7778 }, { "epoch": 0.9895687571555781, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.359296798706055, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8691371083259583, "num_tokens": 296737649.0, "step": 7779 }, { "epoch": 0.9896959674341687, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.383111000061035, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8659502267837524, "num_tokens": 296776915.0, "step": 7780 }, { "epoch": 0.9898231777127592, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.350669860839844, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8698825240135193, "num_tokens": 296811491.0, "step": 7781 }, { "epoch": 0.9899503879913497, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.428661346435547, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8679437637329102, "num_tokens": 296847602.0, "step": 7782 }, { "epoch": 0.9900775982699402, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.400314331054688, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8477163314819336, "num_tokens": 296884737.0, "step": 7783 }, { "epoch": 0.9902048085485308, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.391807556152344, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8514564633369446, "num_tokens": 296927780.0, "step": 7784 }, { "epoch": 0.9903320188271212, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.403213500976562, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.878130316734314, "num_tokens": 296964200.0, "step": 7785 }, { "epoch": 0.9904592291057117, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.391427993774414, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8549938797950745, "num_tokens": 297009927.0, "step": 7786 }, { "epoch": 0.9905864393843022, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.341609954833984, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8519859313964844, "num_tokens": 297050788.0, "step": 7787 }, { "epoch": 0.9907136496628928, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.34656810760498, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8684785962104797, "num_tokens": 297092398.0, "step": 7788 }, { "epoch": 0.9908408599414833, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.431013107299805, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8617059588432312, "num_tokens": 297130728.0, "step": 7789 }, { "epoch": 0.9909680702200738, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.334197998046875, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8866682648658752, "num_tokens": 297169124.0, "step": 7790 }, { "epoch": 0.9910952804986642, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.41179084777832, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8727318644523621, "num_tokens": 297204914.0, "step": 7791 }, { "epoch": 0.9912224907772548, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.453815460205078, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8604938387870789, "num_tokens": 297240551.0, "step": 7792 }, { "epoch": 0.9913497010558453, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.333879470825195, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.855154275894165, "num_tokens": 297287160.0, "step": 7793 }, { "epoch": 0.9914769113344358, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.460365295410156, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8657571077346802, "num_tokens": 297323394.0, "step": 7794 }, { "epoch": 0.9916041216130264, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.439687728881836, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8421156406402588, "num_tokens": 297357808.0, "step": 7795 }, { "epoch": 0.9917313318916169, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.354427337646484, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8593076467514038, "num_tokens": 297397300.0, "step": 7796 }, { "epoch": 0.9918585421702073, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.374298095703125, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8726745247840881, "num_tokens": 297434699.0, "step": 7797 }, { "epoch": 0.9919857524487978, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.32100772857666, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.862897515296936, "num_tokens": 297473939.0, "step": 7798 }, { "epoch": 0.9921129627273884, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.483789443969727, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8561687469482422, "num_tokens": 297511154.0, "step": 7799 }, { "epoch": 0.9922401730059789, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.33242416381836, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.855116069316864, "num_tokens": 297556092.0, "step": 7800 }, { "epoch": 0.9923673832845694, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.397275924682617, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8680505752563477, "num_tokens": 297598160.0, "step": 7801 }, { "epoch": 0.9924945935631599, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.393136024475098, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8803709745407104, "num_tokens": 297638706.0, "step": 7802 }, { "epoch": 0.9926218038417504, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.399836540222168, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8682772517204285, "num_tokens": 297675955.0, "step": 7803 }, { "epoch": 0.9927490141203409, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.376585960388184, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8586529493331909, "num_tokens": 297715362.0, "step": 7804 }, { "epoch": 0.9928762243989314, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.47376537322998, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8493726849555969, "num_tokens": 297750931.0, "step": 7805 }, { "epoch": 0.993003434677522, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.387262344360352, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8557875156402588, "num_tokens": 297786107.0, "step": 7806 }, { "epoch": 0.9931306449561125, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.380570411682129, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8636058568954468, "num_tokens": 297824782.0, "step": 7807 }, { "epoch": 0.993257855234703, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.422640800476074, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8594979047775269, "num_tokens": 297861748.0, "step": 7808 }, { "epoch": 0.9933850655132934, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.351127624511719, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8634757995605469, "num_tokens": 297902722.0, "step": 7809 }, { "epoch": 0.993512275791884, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.380699157714844, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8651341795921326, "num_tokens": 297948709.0, "step": 7810 }, { "epoch": 0.9936394860704745, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.360466003417969, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8660649657249451, "num_tokens": 297986446.0, "step": 7811 }, { "epoch": 0.993766696349065, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.429715156555176, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8557316064834595, "num_tokens": 298024004.0, "step": 7812 }, { "epoch": 0.9938939066276555, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.390990257263184, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8541812896728516, "num_tokens": 298065518.0, "step": 7813 }, { "epoch": 0.9940211169062461, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.34537124633789, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8563357591629028, "num_tokens": 298108138.0, "step": 7814 }, { "epoch": 0.9941483271848366, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.346932411193848, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8636279106140137, "num_tokens": 298149348.0, "step": 7815 }, { "epoch": 0.994275537463427, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.407130241394043, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8611162900924683, "num_tokens": 298186347.0, "step": 7816 }, { "epoch": 0.9944027477420175, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.346510887145996, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8429815173149109, "num_tokens": 298223322.0, "step": 7817 }, { "epoch": 0.9945299580206081, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.4249849319458, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8590901494026184, "num_tokens": 298258794.0, "step": 7818 }, { "epoch": 0.9946571682991986, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.405576705932617, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8568407297134399, "num_tokens": 298297510.0, "step": 7819 }, { "epoch": 0.9947843785777891, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.41001033782959, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8676986694335938, "num_tokens": 298333784.0, "step": 7820 }, { "epoch": 0.9949115888563796, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.392845153808594, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8699209690093994, "num_tokens": 298371241.0, "step": 7821 }, { "epoch": 0.9950387991349701, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.378437995910645, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8492769002914429, "num_tokens": 298407902.0, "step": 7822 }, { "epoch": 0.9951660094135606, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.380071640014648, "learning_rate": 1e-06, "loss": 0.5015, "mean_token_accuracy": 0.8392095565795898, "num_tokens": 298452339.0, "step": 7823 }, { "epoch": 0.9952932196921511, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.379912376403809, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8538150787353516, "num_tokens": 298495854.0, "step": 7824 }, { "epoch": 0.9954204299707416, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.419798851013184, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8571820259094238, "num_tokens": 298529642.0, "step": 7825 }, { "epoch": 0.9955476402493322, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.358317375183105, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8642069101333618, "num_tokens": 298574396.0, "step": 7826 }, { "epoch": 0.9956748505279227, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.440613746643066, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8496690988540649, "num_tokens": 298612457.0, "step": 7827 }, { "epoch": 0.9958020608065131, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.396805763244629, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8627867102622986, "num_tokens": 298645061.0, "step": 7828 }, { "epoch": 0.9959292710851037, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.390580177307129, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8606407642364502, "num_tokens": 298688014.0, "step": 7829 }, { "epoch": 0.9960564813636942, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.387222290039062, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8657280802726746, "num_tokens": 298727815.0, "step": 7830 }, { "epoch": 0.9961836916422847, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.380961418151855, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8844766616821289, "num_tokens": 298765663.0, "step": 7831 }, { "epoch": 0.9963109019208752, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.355653762817383, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8673446178436279, "num_tokens": 298799360.0, "step": 7832 }, { "epoch": 0.9964381121994658, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.385539054870605, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.862451434135437, "num_tokens": 298835669.0, "step": 7833 }, { "epoch": 0.9965653224780562, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.343245506286621, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.861290693283081, "num_tokens": 298866871.0, "step": 7834 }, { "epoch": 0.9966925327566467, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.422956466674805, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8547003269195557, "num_tokens": 298904136.0, "step": 7835 }, { "epoch": 0.9968197430352372, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.357439994812012, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8751494884490967, "num_tokens": 298937069.0, "step": 7836 }, { "epoch": 0.9969469533138278, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.3535795211792, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8550771474838257, "num_tokens": 298976778.0, "step": 7837 }, { "epoch": 0.9970741635924183, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.386236190795898, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8732856512069702, "num_tokens": 299010808.0, "step": 7838 }, { "epoch": 0.9972013738710088, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.396941184997559, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8522734045982361, "num_tokens": 299047611.0, "step": 7839 }, { "epoch": 0.9973285841495992, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.409642219543457, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8676284551620483, "num_tokens": 299088668.0, "step": 7840 }, { "epoch": 0.9974557944281898, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.365958213806152, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.8428167104721069, "num_tokens": 299126572.0, "step": 7841 }, { "epoch": 0.9975830047067803, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.41635799407959, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8743651509284973, "num_tokens": 299164806.0, "step": 7842 }, { "epoch": 0.9977102149853708, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.347124099731445, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.858627200126648, "num_tokens": 299207039.0, "step": 7843 }, { "epoch": 0.9978374252639614, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.422163009643555, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.871761679649353, "num_tokens": 299243242.0, "step": 7844 }, { "epoch": 0.9979646355425519, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.447253227233887, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.869033932685852, "num_tokens": 299282326.0, "step": 7845 }, { "epoch": 0.9980918458211423, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.356351852416992, "learning_rate": 1e-06, "loss": 0.331, "mean_token_accuracy": 0.8914449214935303, "num_tokens": 299316645.0, "step": 7846 }, { "epoch": 0.9982190560997328, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.519485473632812, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8652526140213013, "num_tokens": 299349932.0, "step": 7847 }, { "epoch": 0.9983462663783234, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.455061912536621, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8414900898933411, "num_tokens": 299392452.0, "step": 7848 }, { "epoch": 0.9984734766569139, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.40302848815918, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8682511448860168, "num_tokens": 299431479.0, "step": 7849 }, { "epoch": 0.9986006869355044, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.37838363647461, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8609142899513245, "num_tokens": 299475585.0, "step": 7850 }, { "epoch": 0.9987278972140949, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.442444801330566, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8790804743766785, "num_tokens": 299514026.0, "step": 7851 }, { "epoch": 0.9988551074926854, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.46221923828125, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8526078462600708, "num_tokens": 299550858.0, "step": 7852 }, { "epoch": 0.9989823177712759, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.368782043457031, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8564164638519287, "num_tokens": 299586831.0, "step": 7853 }, { "epoch": 0.9991095280498664, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.419768333435059, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8728406429290771, "num_tokens": 299623116.0, "step": 7854 }, { "epoch": 0.9992367383284569, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.35036563873291, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8660717010498047, "num_tokens": 299662165.0, "step": 7855 }, { "epoch": 0.9993639486070475, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.367940902709961, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8601977825164795, "num_tokens": 299702581.0, "step": 7856 }, { "epoch": 0.999491158885638, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.451738357543945, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8487577438354492, "num_tokens": 299742559.0, "step": 7857 }, { "epoch": 0.9996183691642284, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.440897941589355, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8788771629333496, "num_tokens": 299778637.0, "step": 7858 }, { "epoch": 0.9997455794428189, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.40998363494873, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8721170425415039, "num_tokens": 299812808.0, "step": 7859 }, { "epoch": 0.9998727897214095, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.432544708251953, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8599804639816284, "num_tokens": 299848987.0, "step": 7860 }, { "epoch": 1.0, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.376662254333496, "learning_rate": 1e-06, "loss": 0.4972, "mean_token_accuracy": 0.8451052904129028, "num_tokens": 299886286.0, "step": 7861 }, { "epoch": 1.0001272102785905, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.428640365600586, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8585878610610962, "num_tokens": 299925456.0, "step": 7862 }, { "epoch": 1.000254420557181, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.405359268188477, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8613710403442383, "num_tokens": 299965936.0, "step": 7863 }, { "epoch": 1.0003816308357716, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.393718719482422, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8773688077926636, "num_tokens": 300003181.0, "step": 7864 }, { "epoch": 1.0005088411143621, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.447853088378906, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8624810576438904, "num_tokens": 300040502.0, "step": 7865 }, { "epoch": 1.0006360513929526, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.358616828918457, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8799353837966919, "num_tokens": 300078795.0, "step": 7866 }, { "epoch": 1.0007632616715432, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.411474227905273, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.874896764755249, "num_tokens": 300115188.0, "step": 7867 }, { "epoch": 1.0008904719501335, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.31716251373291, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.86680668592453, "num_tokens": 300158583.0, "step": 7868 }, { "epoch": 1.001017682228724, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.42739200592041, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8700689077377319, "num_tokens": 300194415.0, "step": 7869 }, { "epoch": 1.0011448925073145, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.429506301879883, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8607525825500488, "num_tokens": 300235066.0, "step": 7870 }, { "epoch": 1.001272102785905, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.382147789001465, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8672351241111755, "num_tokens": 300272167.0, "step": 7871 }, { "epoch": 1.0013993130644956, "ewc_loss": 0.0234375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.506665229797363, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8562732934951782, "num_tokens": 300306032.0, "step": 7872 }, { "epoch": 1.0015265233430861, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.443641662597656, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8624743223190308, "num_tokens": 300338433.0, "step": 7873 }, { "epoch": 1.0016537336216766, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.447647094726562, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8548572063446045, "num_tokens": 300376527.0, "step": 7874 }, { "epoch": 1.0017809439002672, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.456880569458008, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.882029116153717, "num_tokens": 300417309.0, "step": 7875 }, { "epoch": 1.0019081541788577, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.385381698608398, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8662364482879639, "num_tokens": 300459438.0, "step": 7876 }, { "epoch": 1.0020353644574482, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.403913497924805, "learning_rate": 1e-06, "loss": 0.4898, "mean_token_accuracy": 0.8416451811790466, "num_tokens": 300501443.0, "step": 7877 }, { "epoch": 1.0021625747360388, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.487810134887695, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8730460405349731, "num_tokens": 300531918.0, "step": 7878 }, { "epoch": 1.0022897850146293, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.366132736206055, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8593116998672485, "num_tokens": 300575084.0, "step": 7879 }, { "epoch": 1.0024169952932196, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.447032928466797, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8659825325012207, "num_tokens": 300608737.0, "step": 7880 }, { "epoch": 1.0025442055718101, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.384937286376953, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8534802198410034, "num_tokens": 300647855.0, "step": 7881 }, { "epoch": 1.0026714158504006, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.430685043334961, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.866531252861023, "num_tokens": 300689763.0, "step": 7882 }, { "epoch": 1.0027986261289912, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.443807601928711, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8758789300918579, "num_tokens": 300725838.0, "step": 7883 }, { "epoch": 1.0029258364075817, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.397439956665039, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.871458113193512, "num_tokens": 300760423.0, "step": 7884 }, { "epoch": 1.0030530466861722, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.460223197937012, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8713470697402954, "num_tokens": 300796236.0, "step": 7885 }, { "epoch": 1.0031802569647628, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.37330436706543, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8703389167785645, "num_tokens": 300835213.0, "step": 7886 }, { "epoch": 1.0033074672433533, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.421104431152344, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8725365996360779, "num_tokens": 300873363.0, "step": 7887 }, { "epoch": 1.0034346775219438, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.459372520446777, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8607984781265259, "num_tokens": 300910233.0, "step": 7888 }, { "epoch": 1.0035618878005343, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.388206481933594, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8728575706481934, "num_tokens": 300946986.0, "step": 7889 }, { "epoch": 1.0036890980791249, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.410714149475098, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.867999792098999, "num_tokens": 300985750.0, "step": 7890 }, { "epoch": 1.0038163083577154, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.365130424499512, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8646156787872314, "num_tokens": 301032820.0, "step": 7891 }, { "epoch": 1.0039435186363057, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.4007568359375, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8559641242027283, "num_tokens": 301070825.0, "step": 7892 }, { "epoch": 1.0040707289148962, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.441137313842773, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.861751914024353, "num_tokens": 301110764.0, "step": 7893 }, { "epoch": 1.0041979391934868, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.33192253112793, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8680251240730286, "num_tokens": 301152421.0, "step": 7894 }, { "epoch": 1.0043251494720773, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.449026107788086, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8663070797920227, "num_tokens": 301192780.0, "step": 7895 }, { "epoch": 1.0044523597506678, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.412301063537598, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8621271848678589, "num_tokens": 301228147.0, "step": 7896 }, { "epoch": 1.0045795700292584, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.434097290039062, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8752619624137878, "num_tokens": 301259568.0, "step": 7897 }, { "epoch": 1.0047067803078489, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.38604736328125, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8668453097343445, "num_tokens": 301296483.0, "step": 7898 }, { "epoch": 1.0048339905864394, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.370519638061523, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8688995838165283, "num_tokens": 301337437.0, "step": 7899 }, { "epoch": 1.00496120086503, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.504855155944824, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.864487886428833, "num_tokens": 301368777.0, "step": 7900 }, { "epoch": 1.0050884111436205, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.42435073852539, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8625396490097046, "num_tokens": 301404885.0, "step": 7901 }, { "epoch": 1.005215621422211, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.353553771972656, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8556978702545166, "num_tokens": 301445308.0, "step": 7902 }, { "epoch": 1.0053428317008015, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.504292488098145, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.862261176109314, "num_tokens": 301484845.0, "step": 7903 }, { "epoch": 1.0054700419793918, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.42380428314209, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8629531264305115, "num_tokens": 301524498.0, "step": 7904 }, { "epoch": 1.0055972522579824, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.41911792755127, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8710764646530151, "num_tokens": 301565522.0, "step": 7905 }, { "epoch": 1.0057244625365729, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.348870277404785, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8613831996917725, "num_tokens": 301603724.0, "step": 7906 }, { "epoch": 1.0058516728151634, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.519569396972656, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.862341046333313, "num_tokens": 301636650.0, "step": 7907 }, { "epoch": 1.005978883093754, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.40957260131836, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8557996153831482, "num_tokens": 301675328.0, "step": 7908 }, { "epoch": 1.0061060933723445, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.46915340423584, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8589062690734863, "num_tokens": 301710058.0, "step": 7909 }, { "epoch": 1.006233303650935, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.384029388427734, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8595561981201172, "num_tokens": 301753567.0, "step": 7910 }, { "epoch": 1.0063605139295255, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.378812789916992, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8724664449691772, "num_tokens": 301792895.0, "step": 7911 }, { "epoch": 1.006487724208116, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.431334495544434, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8761712312698364, "num_tokens": 301833001.0, "step": 7912 }, { "epoch": 1.0066149344867066, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.432952880859375, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8693061470985413, "num_tokens": 301871623.0, "step": 7913 }, { "epoch": 1.006742144765297, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.445876121520996, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8770892024040222, "num_tokens": 301903750.0, "step": 7914 }, { "epoch": 1.0068693550438876, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.380701065063477, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8664960265159607, "num_tokens": 301943079.0, "step": 7915 }, { "epoch": 1.0069965653224782, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.395783424377441, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8700751066207886, "num_tokens": 301980031.0, "step": 7916 }, { "epoch": 1.0071237756010685, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.46336841583252, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8655027151107788, "num_tokens": 302020454.0, "step": 7917 }, { "epoch": 1.007250985879659, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.366978645324707, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8602977991104126, "num_tokens": 302064391.0, "step": 7918 }, { "epoch": 1.0073781961582495, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.443731307983398, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8667049407958984, "num_tokens": 302101598.0, "step": 7919 }, { "epoch": 1.00750540643684, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.407758712768555, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8695971369743347, "num_tokens": 302138803.0, "step": 7920 }, { "epoch": 1.0076326167154306, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.441957473754883, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8622986674308777, "num_tokens": 302173682.0, "step": 7921 }, { "epoch": 1.0077598269940211, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.412429809570312, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8617194890975952, "num_tokens": 302207390.0, "step": 7922 }, { "epoch": 1.0078870372726116, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.381316184997559, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8709532618522644, "num_tokens": 302244440.0, "step": 7923 }, { "epoch": 1.0080142475512022, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.499659538269043, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8602465987205505, "num_tokens": 302281058.0, "step": 7924 }, { "epoch": 1.0081414578297927, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.343552589416504, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8818293213844299, "num_tokens": 302319586.0, "step": 7925 }, { "epoch": 1.0082686681083832, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.4967679977417, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8599989414215088, "num_tokens": 302361356.0, "step": 7926 }, { "epoch": 1.0083958783869738, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.393396377563477, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8741822838783264, "num_tokens": 302399974.0, "step": 7927 }, { "epoch": 1.0085230886655643, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.43117618560791, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.867773175239563, "num_tokens": 302435709.0, "step": 7928 }, { "epoch": 1.0086502989441546, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.446028709411621, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8617079854011536, "num_tokens": 302474151.0, "step": 7929 }, { "epoch": 1.0087775092227451, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.50519847869873, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8677000999450684, "num_tokens": 302509877.0, "step": 7930 }, { "epoch": 1.0089047195013356, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.36826229095459, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8787668943405151, "num_tokens": 302547689.0, "step": 7931 }, { "epoch": 1.0090319297799262, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.438186645507812, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8632933497428894, "num_tokens": 302590728.0, "step": 7932 }, { "epoch": 1.0091591400585167, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.463728904724121, "learning_rate": 1e-06, "loss": 0.4907, "mean_token_accuracy": 0.8434272408485413, "num_tokens": 302625709.0, "step": 7933 }, { "epoch": 1.0092863503371072, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.461750984191895, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8736704587936401, "num_tokens": 302662086.0, "step": 7934 }, { "epoch": 1.0094135606156978, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.361824035644531, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8825184106826782, "num_tokens": 302703680.0, "step": 7935 }, { "epoch": 1.0095407708942883, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.447196960449219, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8509888052940369, "num_tokens": 302740426.0, "step": 7936 }, { "epoch": 1.0096679811728788, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.438437461853027, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8783186674118042, "num_tokens": 302780778.0, "step": 7937 }, { "epoch": 1.0097951914514693, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.41498851776123, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8690527081489563, "num_tokens": 302816012.0, "step": 7938 }, { "epoch": 1.0099224017300599, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.456911087036133, "learning_rate": 1e-06, "loss": 0.4872, "mean_token_accuracy": 0.8442569971084595, "num_tokens": 302858325.0, "step": 7939 }, { "epoch": 1.0100496120086504, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.44821834564209, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8663246035575867, "num_tokens": 302898210.0, "step": 7940 }, { "epoch": 1.0101768222872407, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.45586109161377, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8660577535629272, "num_tokens": 302936072.0, "step": 7941 }, { "epoch": 1.0103040325658312, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.459135055541992, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8578959703445435, "num_tokens": 302978941.0, "step": 7942 }, { "epoch": 1.0104312428444218, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.472247123718262, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8583837747573853, "num_tokens": 303016664.0, "step": 7943 }, { "epoch": 1.0105584531230123, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.462347984313965, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8616020679473877, "num_tokens": 303063313.0, "step": 7944 }, { "epoch": 1.0106856634016028, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.468164443969727, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8557233214378357, "num_tokens": 303103786.0, "step": 7945 }, { "epoch": 1.0108128736801933, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.463851928710938, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8711796998977661, "num_tokens": 303136928.0, "step": 7946 }, { "epoch": 1.0109400839587839, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.433915138244629, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8734034299850464, "num_tokens": 303174854.0, "step": 7947 }, { "epoch": 1.0110672942373744, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.380058288574219, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8634976148605347, "num_tokens": 303214234.0, "step": 7948 }, { "epoch": 1.011194504515965, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.38611125946045, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8790037035942078, "num_tokens": 303251830.0, "step": 7949 }, { "epoch": 1.0113217147945555, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.398555755615234, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8553363084793091, "num_tokens": 303291564.0, "step": 7950 }, { "epoch": 1.011448925073146, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.495160102844238, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8629539608955383, "num_tokens": 303328642.0, "step": 7951 }, { "epoch": 1.0115761353517365, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.414168357849121, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8635058403015137, "num_tokens": 303373693.0, "step": 7952 }, { "epoch": 1.0117033456303268, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.463069915771484, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.869649350643158, "num_tokens": 303407124.0, "step": 7953 }, { "epoch": 1.0118305559089174, "ewc_loss": 0.023681640625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.437756538391113, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8762958645820618, "num_tokens": 303441586.0, "step": 7954 }, { "epoch": 1.0119577661875079, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.475136756896973, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8668273687362671, "num_tokens": 303473593.0, "step": 7955 }, { "epoch": 1.0120849764660984, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.389161109924316, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8620652556419373, "num_tokens": 303510239.0, "step": 7956 }, { "epoch": 1.012212186744689, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.481796264648438, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8627550601959229, "num_tokens": 303544709.0, "step": 7957 }, { "epoch": 1.0123393970232795, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.409010887145996, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8748539686203003, "num_tokens": 303581990.0, "step": 7958 }, { "epoch": 1.01246660730187, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.452140808105469, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8690041899681091, "num_tokens": 303625167.0, "step": 7959 }, { "epoch": 1.0125938175804605, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.411977767944336, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8662481307983398, "num_tokens": 303664751.0, "step": 7960 }, { "epoch": 1.012721027859051, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.44533634185791, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8634108304977417, "num_tokens": 303702190.0, "step": 7961 }, { "epoch": 1.0128482381376416, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.39159107208252, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8797566890716553, "num_tokens": 303742872.0, "step": 7962 }, { "epoch": 1.012975448416232, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.3784761428833, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8580037355422974, "num_tokens": 303781265.0, "step": 7963 }, { "epoch": 1.0131026586948226, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.430612564086914, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8663249015808105, "num_tokens": 303820686.0, "step": 7964 }, { "epoch": 1.0132298689734132, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.386761665344238, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8682968616485596, "num_tokens": 303863428.0, "step": 7965 }, { "epoch": 1.0133570792520035, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.418973922729492, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8609788417816162, "num_tokens": 303903847.0, "step": 7966 }, { "epoch": 1.013484289530594, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.38766098022461, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8574650287628174, "num_tokens": 303943083.0, "step": 7967 }, { "epoch": 1.0136114998091845, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.399273872375488, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8762339353561401, "num_tokens": 303978773.0, "step": 7968 }, { "epoch": 1.013738710087775, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.367415428161621, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8724469542503357, "num_tokens": 304017058.0, "step": 7969 }, { "epoch": 1.0138659203663656, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.406139373779297, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8569732904434204, "num_tokens": 304054104.0, "step": 7970 }, { "epoch": 1.013993130644956, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.437232971191406, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8795127868652344, "num_tokens": 304091100.0, "step": 7971 }, { "epoch": 1.0141203409235466, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.430098533630371, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8666424751281738, "num_tokens": 304127747.0, "step": 7972 }, { "epoch": 1.0142475512021372, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.406047821044922, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8671172857284546, "num_tokens": 304171089.0, "step": 7973 }, { "epoch": 1.0143747614807277, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.454005241394043, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.856728196144104, "num_tokens": 304210207.0, "step": 7974 }, { "epoch": 1.0145019717593182, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.434532165527344, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8828645348548889, "num_tokens": 304249423.0, "step": 7975 }, { "epoch": 1.0146291820379088, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.462913513183594, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.870317280292511, "num_tokens": 304286872.0, "step": 7976 }, { "epoch": 1.0147563923164993, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.425593376159668, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8699730634689331, "num_tokens": 304323033.0, "step": 7977 }, { "epoch": 1.0148836025950896, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.485695838928223, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8692586421966553, "num_tokens": 304360096.0, "step": 7978 }, { "epoch": 1.0150108128736801, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.344852447509766, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8861937522888184, "num_tokens": 304396364.0, "step": 7979 }, { "epoch": 1.0151380231522706, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.504154205322266, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8671820163726807, "num_tokens": 304431583.0, "step": 7980 }, { "epoch": 1.0152652334308612, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.384743690490723, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8601418733596802, "num_tokens": 304470531.0, "step": 7981 }, { "epoch": 1.0153924437094517, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.494998931884766, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8701692819595337, "num_tokens": 304505187.0, "step": 7982 }, { "epoch": 1.0155196539880422, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.431180000305176, "learning_rate": 1e-06, "loss": 0.4848, "mean_token_accuracy": 0.8445439338684082, "num_tokens": 304542685.0, "step": 7983 }, { "epoch": 1.0156468642666328, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.430952072143555, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.866783857345581, "num_tokens": 304579204.0, "step": 7984 }, { "epoch": 1.0157740745452233, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.427727699279785, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8735936284065247, "num_tokens": 304615224.0, "step": 7985 }, { "epoch": 1.0159012848238138, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.492975234985352, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8615940809249878, "num_tokens": 304653513.0, "step": 7986 }, { "epoch": 1.0160284951024043, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.403016090393066, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8724830150604248, "num_tokens": 304686744.0, "step": 7987 }, { "epoch": 1.0161557053809949, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.465576171875, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8656721711158752, "num_tokens": 304727262.0, "step": 7988 }, { "epoch": 1.0162829156595854, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.479338645935059, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8641292452812195, "num_tokens": 304763418.0, "step": 7989 }, { "epoch": 1.0164101259381757, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.449460983276367, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.87010657787323, "num_tokens": 304800763.0, "step": 7990 }, { "epoch": 1.0165373362167662, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.456053733825684, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.8845659494400024, "num_tokens": 304833136.0, "step": 7991 }, { "epoch": 1.0166645464953568, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.47242259979248, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8762943744659424, "num_tokens": 304873208.0, "step": 7992 }, { "epoch": 1.0167917567739473, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.489232063293457, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8673781156539917, "num_tokens": 304914237.0, "step": 7993 }, { "epoch": 1.0169189670525378, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.400675773620605, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8823181986808777, "num_tokens": 304955453.0, "step": 7994 }, { "epoch": 1.0170461773311283, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.508974075317383, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8664051294326782, "num_tokens": 304990533.0, "step": 7995 }, { "epoch": 1.0171733876097189, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.464394569396973, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8651666641235352, "num_tokens": 305024849.0, "step": 7996 }, { "epoch": 1.0173005978883094, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.446915626525879, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.863404393196106, "num_tokens": 305061615.0, "step": 7997 }, { "epoch": 1.0174278081669, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.43004035949707, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.847400426864624, "num_tokens": 305102251.0, "step": 7998 }, { "epoch": 1.0175550184454905, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.439236640930176, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8751817941665649, "num_tokens": 305135405.0, "step": 7999 }, { "epoch": 1.017682228724081, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.386439323425293, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8655979633331299, "num_tokens": 305169005.0, "step": 8000 }, { "epoch": 1.0178094390026715, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.463679313659668, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8620944619178772, "num_tokens": 305209050.0, "step": 8001 }, { "epoch": 1.0179366492812618, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.47089958190918, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8664168119430542, "num_tokens": 305238250.0, "step": 8002 }, { "epoch": 1.0180638595598523, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.449970245361328, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.86937016248703, "num_tokens": 305283544.0, "step": 8003 }, { "epoch": 1.0181910698384429, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.466130256652832, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8505980968475342, "num_tokens": 305320548.0, "step": 8004 }, { "epoch": 1.0183182801170334, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.423723220825195, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8731430172920227, "num_tokens": 305359560.0, "step": 8005 }, { "epoch": 1.018445490395624, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.45304012298584, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8634560108184814, "num_tokens": 305396680.0, "step": 8006 }, { "epoch": 1.0185727006742145, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.470783233642578, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8766022324562073, "num_tokens": 305432856.0, "step": 8007 }, { "epoch": 1.018699910952805, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.448724746704102, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8681708574295044, "num_tokens": 305469973.0, "step": 8008 }, { "epoch": 1.0188271212313955, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.48633098602295, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8726947903633118, "num_tokens": 305505576.0, "step": 8009 }, { "epoch": 1.018954331509986, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.482343673706055, "learning_rate": 1e-06, "loss": 0.495, "mean_token_accuracy": 0.8487647175788879, "num_tokens": 305542037.0, "step": 8010 }, { "epoch": 1.0190815417885766, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.42574405670166, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8734868764877319, "num_tokens": 305587784.0, "step": 8011 }, { "epoch": 1.019208752067167, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.490030288696289, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8596750497817993, "num_tokens": 305621888.0, "step": 8012 }, { "epoch": 1.0193359623457576, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.451717376708984, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8862910866737366, "num_tokens": 305663844.0, "step": 8013 }, { "epoch": 1.0194631726243482, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.432663917541504, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8812482357025146, "num_tokens": 305695913.0, "step": 8014 }, { "epoch": 1.0195903829029385, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.421526908874512, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8651913404464722, "num_tokens": 305733632.0, "step": 8015 }, { "epoch": 1.019717593181529, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.479851722717285, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8505515456199646, "num_tokens": 305766917.0, "step": 8016 }, { "epoch": 1.0198448034601195, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.389575958251953, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8614315986633301, "num_tokens": 305805768.0, "step": 8017 }, { "epoch": 1.01997201373871, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.424528121948242, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8674139380455017, "num_tokens": 305843412.0, "step": 8018 }, { "epoch": 1.0200992240173006, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.384629249572754, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8815034627914429, "num_tokens": 305880523.0, "step": 8019 }, { "epoch": 1.020226434295891, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.42750358581543, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8691446781158447, "num_tokens": 305915379.0, "step": 8020 }, { "epoch": 1.0203536445744816, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.426180839538574, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8806781768798828, "num_tokens": 305952471.0, "step": 8021 }, { "epoch": 1.0204808548530722, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.387730598449707, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8661516904830933, "num_tokens": 305994023.0, "step": 8022 }, { "epoch": 1.0206080651316627, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.454391479492188, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8716011643409729, "num_tokens": 306026563.0, "step": 8023 }, { "epoch": 1.0207352754102532, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.444250106811523, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8635009527206421, "num_tokens": 306067913.0, "step": 8024 }, { "epoch": 1.0208624856888437, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.407325744628906, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8646537065505981, "num_tokens": 306108692.0, "step": 8025 }, { "epoch": 1.0209896959674343, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.423989295959473, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8676114082336426, "num_tokens": 306148082.0, "step": 8026 }, { "epoch": 1.0211169062460246, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.440125465393066, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8745525479316711, "num_tokens": 306182800.0, "step": 8027 }, { "epoch": 1.021244116524615, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.386174201965332, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8639948964118958, "num_tokens": 306226741.0, "step": 8028 }, { "epoch": 1.0213713268032056, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.457362174987793, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8652678728103638, "num_tokens": 306265993.0, "step": 8029 }, { "epoch": 1.0214985370817962, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.450486183166504, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8706722855567932, "num_tokens": 306306511.0, "step": 8030 }, { "epoch": 1.0216257473603867, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.391488075256348, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8811643123626709, "num_tokens": 306337687.0, "step": 8031 }, { "epoch": 1.0217529576389772, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.420594215393066, "learning_rate": 1e-06, "loss": 0.5134, "mean_token_accuracy": 0.83246910572052, "num_tokens": 306379646.0, "step": 8032 }, { "epoch": 1.0218801679175677, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.517717361450195, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8669872879981995, "num_tokens": 306415389.0, "step": 8033 }, { "epoch": 1.0220073781961583, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.417196273803711, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8620342016220093, "num_tokens": 306452035.0, "step": 8034 }, { "epoch": 1.0221345884747488, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.434494018554688, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8667377233505249, "num_tokens": 306496277.0, "step": 8035 }, { "epoch": 1.0222617987533393, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.479741096496582, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8719560503959656, "num_tokens": 306528212.0, "step": 8036 }, { "epoch": 1.0223890090319299, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.501254081726074, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8707238435745239, "num_tokens": 306562047.0, "step": 8037 }, { "epoch": 1.0225162193105204, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.494699478149414, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8566502332687378, "num_tokens": 306601298.0, "step": 8038 }, { "epoch": 1.0226434295891107, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.432279586791992, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8670002222061157, "num_tokens": 306640961.0, "step": 8039 }, { "epoch": 1.0227706398677012, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.467887878417969, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8849508762359619, "num_tokens": 306679507.0, "step": 8040 }, { "epoch": 1.0228978501462918, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.487353324890137, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8577311635017395, "num_tokens": 306717337.0, "step": 8041 }, { "epoch": 1.0230250604248823, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.467737197875977, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.841344952583313, "num_tokens": 306754152.0, "step": 8042 }, { "epoch": 1.0231522707034728, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.471308708190918, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8599995374679565, "num_tokens": 306794887.0, "step": 8043 }, { "epoch": 1.0232794809820633, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.458024978637695, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8697364330291748, "num_tokens": 306828374.0, "step": 8044 }, { "epoch": 1.0234066912606539, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.447542190551758, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8704333305358887, "num_tokens": 306865285.0, "step": 8045 }, { "epoch": 1.0235339015392444, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.463216781616211, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.872979998588562, "num_tokens": 306897195.0, "step": 8046 }, { "epoch": 1.023661111817835, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.463508605957031, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8654894828796387, "num_tokens": 306935352.0, "step": 8047 }, { "epoch": 1.0237883220964255, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.495948791503906, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8725554943084717, "num_tokens": 306964621.0, "step": 8048 }, { "epoch": 1.023915532375016, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.425816535949707, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8658145070075989, "num_tokens": 307003451.0, "step": 8049 }, { "epoch": 1.0240427426536065, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.54694652557373, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8664052486419678, "num_tokens": 307044695.0, "step": 8050 }, { "epoch": 1.0241699529321968, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.461066246032715, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8736090064048767, "num_tokens": 307080603.0, "step": 8051 }, { "epoch": 1.0242971632107873, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.452191352844238, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8606959581375122, "num_tokens": 307122300.0, "step": 8052 }, { "epoch": 1.0244243734893779, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.515951156616211, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8727648258209229, "num_tokens": 307158685.0, "step": 8053 }, { "epoch": 1.0245515837679684, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.477737426757812, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8728394508361816, "num_tokens": 307199104.0, "step": 8054 }, { "epoch": 1.024678794046559, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.56297492980957, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.859123945236206, "num_tokens": 307235840.0, "step": 8055 }, { "epoch": 1.0248060043251495, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.466148376464844, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.855941653251648, "num_tokens": 307273089.0, "step": 8056 }, { "epoch": 1.02493321460374, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.470666885375977, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8674020171165466, "num_tokens": 307318825.0, "step": 8057 }, { "epoch": 1.0250604248823305, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.444685935974121, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8730459809303284, "num_tokens": 307357006.0, "step": 8058 }, { "epoch": 1.025187635160921, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.499823570251465, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8649968504905701, "num_tokens": 307390274.0, "step": 8059 }, { "epoch": 1.0253148454395116, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.448969841003418, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8747515082359314, "num_tokens": 307427157.0, "step": 8060 }, { "epoch": 1.025442055718102, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.564026832580566, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8643999099731445, "num_tokens": 307466555.0, "step": 8061 }, { "epoch": 1.0255692659966926, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.450263977050781, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8591734170913696, "num_tokens": 307503375.0, "step": 8062 }, { "epoch": 1.0256964762752832, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.484010696411133, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8761212229728699, "num_tokens": 307543170.0, "step": 8063 }, { "epoch": 1.0258236865538735, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.476408004760742, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8633034229278564, "num_tokens": 307582890.0, "step": 8064 }, { "epoch": 1.025950896832464, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.524690628051758, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8588359951972961, "num_tokens": 307617643.0, "step": 8065 }, { "epoch": 1.0260781071110545, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.507145881652832, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8635212779045105, "num_tokens": 307652258.0, "step": 8066 }, { "epoch": 1.026205317389645, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.47481918334961, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.862278163433075, "num_tokens": 307693245.0, "step": 8067 }, { "epoch": 1.0263325276682356, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.468456268310547, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.869060754776001, "num_tokens": 307728751.0, "step": 8068 }, { "epoch": 1.026459737946826, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.497380256652832, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8553696870803833, "num_tokens": 307767869.0, "step": 8069 }, { "epoch": 1.0265869482254166, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.51539134979248, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8755956888198853, "num_tokens": 307809438.0, "step": 8070 }, { "epoch": 1.0267141585040072, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.48176383972168, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8469371199607849, "num_tokens": 307848752.0, "step": 8071 }, { "epoch": 1.0268413687825977, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.507780075073242, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8614291548728943, "num_tokens": 307886080.0, "step": 8072 }, { "epoch": 1.0269685790611882, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.466938018798828, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.880028247833252, "num_tokens": 307922854.0, "step": 8073 }, { "epoch": 1.0270957893397787, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.495379447937012, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.864141583442688, "num_tokens": 307960032.0, "step": 8074 }, { "epoch": 1.0272229996183693, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.50505542755127, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8658295273780823, "num_tokens": 307996074.0, "step": 8075 }, { "epoch": 1.0273502098969596, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.52479362487793, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8788900375366211, "num_tokens": 308030833.0, "step": 8076 }, { "epoch": 1.02747742017555, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.496169090270996, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8766607046127319, "num_tokens": 308079620.0, "step": 8077 }, { "epoch": 1.0276046304541406, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.500123977661133, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8608063459396362, "num_tokens": 308118555.0, "step": 8078 }, { "epoch": 1.0277318407327312, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.523366928100586, "learning_rate": 1e-06, "loss": 0.4931, "mean_token_accuracy": 0.8493030071258545, "num_tokens": 308154410.0, "step": 8079 }, { "epoch": 1.0278590510113217, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.502754211425781, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8663181066513062, "num_tokens": 308191990.0, "step": 8080 }, { "epoch": 1.0279862612899122, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.428162574768066, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.846565842628479, "num_tokens": 308231047.0, "step": 8081 }, { "epoch": 1.0281134715685027, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.501914024353027, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8806666135787964, "num_tokens": 308269487.0, "step": 8082 }, { "epoch": 1.0282406818470933, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.466649055480957, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8663398027420044, "num_tokens": 308304974.0, "step": 8083 }, { "epoch": 1.0283678921256838, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.536214828491211, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.8872362375259399, "num_tokens": 308342072.0, "step": 8084 }, { "epoch": 1.0284951024042743, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.450093269348145, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8628472089767456, "num_tokens": 308374293.0, "step": 8085 }, { "epoch": 1.0286223126828649, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.491484642028809, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8712496757507324, "num_tokens": 308415872.0, "step": 8086 }, { "epoch": 1.0287495229614554, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.478322982788086, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8600587844848633, "num_tokens": 308451198.0, "step": 8087 }, { "epoch": 1.0288767332400457, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.436701774597168, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8738420605659485, "num_tokens": 308490554.0, "step": 8088 }, { "epoch": 1.0290039435186362, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.494672775268555, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.861268162727356, "num_tokens": 308530221.0, "step": 8089 }, { "epoch": 1.0291311537972267, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.467913627624512, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8523242473602295, "num_tokens": 308563634.0, "step": 8090 }, { "epoch": 1.0292583640758173, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.495916366577148, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8666179180145264, "num_tokens": 308605033.0, "step": 8091 }, { "epoch": 1.0293855743544078, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.449376106262207, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8674124479293823, "num_tokens": 308642603.0, "step": 8092 }, { "epoch": 1.0295127846329983, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.543109893798828, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8644592761993408, "num_tokens": 308685189.0, "step": 8093 }, { "epoch": 1.0296399949115889, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.441088676452637, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8666054010391235, "num_tokens": 308726052.0, "step": 8094 }, { "epoch": 1.0297672051901794, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.478856086730957, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8548709154129028, "num_tokens": 308764828.0, "step": 8095 }, { "epoch": 1.02989441546877, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.54189682006836, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8763609528541565, "num_tokens": 308799712.0, "step": 8096 }, { "epoch": 1.0300216257473604, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.503005981445312, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8665966391563416, "num_tokens": 308839879.0, "step": 8097 }, { "epoch": 1.030148836025951, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.514669418334961, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8593242168426514, "num_tokens": 308880362.0, "step": 8098 }, { "epoch": 1.0302760463045415, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.430721282958984, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8763296604156494, "num_tokens": 308921395.0, "step": 8099 }, { "epoch": 1.0304032565831318, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.457483291625977, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.875198245048523, "num_tokens": 308959976.0, "step": 8100 }, { "epoch": 1.0305304668617223, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.513678550720215, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8791476488113403, "num_tokens": 309000702.0, "step": 8101 }, { "epoch": 1.0306576771403129, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.489127159118652, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8845757842063904, "num_tokens": 309036540.0, "step": 8102 }, { "epoch": 1.0307848874189034, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.463902473449707, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8654965758323669, "num_tokens": 309072408.0, "step": 8103 }, { "epoch": 1.030912097697494, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.496254920959473, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8536274433135986, "num_tokens": 309110366.0, "step": 8104 }, { "epoch": 1.0310393079760845, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.515156745910645, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8699740171432495, "num_tokens": 309155194.0, "step": 8105 }, { "epoch": 1.031166518254675, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.469408988952637, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8752681016921997, "num_tokens": 309193747.0, "step": 8106 }, { "epoch": 1.0312937285332655, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.49836540222168, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8544982671737671, "num_tokens": 309234161.0, "step": 8107 }, { "epoch": 1.031420938811856, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.462864875793457, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8555951714515686, "num_tokens": 309272574.0, "step": 8108 }, { "epoch": 1.0315481490904466, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.4591703414917, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8690717220306396, "num_tokens": 309306567.0, "step": 8109 }, { "epoch": 1.031675359369037, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.495170593261719, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.871379017829895, "num_tokens": 309347218.0, "step": 8110 }, { "epoch": 1.0318025696476276, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.456647872924805, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8813276886940002, "num_tokens": 309385701.0, "step": 8111 }, { "epoch": 1.0319297799262181, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.44169807434082, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8646029829978943, "num_tokens": 309424600.0, "step": 8112 }, { "epoch": 1.0320569902048085, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.502781867980957, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8467336893081665, "num_tokens": 309459929.0, "step": 8113 }, { "epoch": 1.032184200483399, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.476998329162598, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8581643104553223, "num_tokens": 309505085.0, "step": 8114 }, { "epoch": 1.0323114107619895, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.491089820861816, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8681498169898987, "num_tokens": 309544746.0, "step": 8115 }, { "epoch": 1.03243862104058, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.484894752502441, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.854677677154541, "num_tokens": 309584301.0, "step": 8116 }, { "epoch": 1.0325658313191706, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.504150390625, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8705167174339294, "num_tokens": 309622570.0, "step": 8117 }, { "epoch": 1.032693041597761, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.500982284545898, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8497506976127625, "num_tokens": 309666417.0, "step": 8118 }, { "epoch": 1.0328202518763516, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.45206069946289, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8569245338439941, "num_tokens": 309705706.0, "step": 8119 }, { "epoch": 1.0329474621549422, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.524158477783203, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8760243654251099, "num_tokens": 309744288.0, "step": 8120 }, { "epoch": 1.0330746724335327, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.44538688659668, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8536567687988281, "num_tokens": 309780406.0, "step": 8121 }, { "epoch": 1.0332018827121232, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.596790313720703, "learning_rate": 1e-06, "loss": 0.5081, "mean_token_accuracy": 0.8349786400794983, "num_tokens": 309811135.0, "step": 8122 }, { "epoch": 1.0333290929907137, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.486028671264648, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8717774748802185, "num_tokens": 309847037.0, "step": 8123 }, { "epoch": 1.0334563032693043, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.418119430541992, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8677480220794678, "num_tokens": 309884274.0, "step": 8124 }, { "epoch": 1.0335835135478946, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.57081413269043, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8789331316947937, "num_tokens": 309915572.0, "step": 8125 }, { "epoch": 1.033710723826485, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.404017448425293, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8712441921234131, "num_tokens": 309954502.0, "step": 8126 }, { "epoch": 1.0338379341050756, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.553403854370117, "learning_rate": 1e-06, "loss": 0.3452, "mean_token_accuracy": 0.8882932662963867, "num_tokens": 309993015.0, "step": 8127 }, { "epoch": 1.0339651443836662, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.505446434020996, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8575395345687866, "num_tokens": 310030840.0, "step": 8128 }, { "epoch": 1.0340923546622567, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.45461654663086, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.867257833480835, "num_tokens": 310067586.0, "step": 8129 }, { "epoch": 1.0342195649408472, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.492112159729004, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8752663135528564, "num_tokens": 310104496.0, "step": 8130 }, { "epoch": 1.0343467752194377, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.519052505493164, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8498265743255615, "num_tokens": 310141162.0, "step": 8131 }, { "epoch": 1.0344739854980283, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.43165397644043, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8641195893287659, "num_tokens": 310182965.0, "step": 8132 }, { "epoch": 1.0346011957766188, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.479752540588379, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8493660092353821, "num_tokens": 310219572.0, "step": 8133 }, { "epoch": 1.0347284060552093, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.531460762023926, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8546706438064575, "num_tokens": 310248711.0, "step": 8134 }, { "epoch": 1.0348556163337999, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.518142700195312, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8501008749008179, "num_tokens": 310287915.0, "step": 8135 }, { "epoch": 1.0349828266123904, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.465791702270508, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.863785982131958, "num_tokens": 310325619.0, "step": 8136 }, { "epoch": 1.0351100368909807, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.503007888793945, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8733481168746948, "num_tokens": 310362456.0, "step": 8137 }, { "epoch": 1.0352372471695712, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.460434913635254, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8610180616378784, "num_tokens": 310399584.0, "step": 8138 }, { "epoch": 1.0353644574481617, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.460780143737793, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8604416847229004, "num_tokens": 310437435.0, "step": 8139 }, { "epoch": 1.0354916677267523, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.473177909851074, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8720587491989136, "num_tokens": 310476320.0, "step": 8140 }, { "epoch": 1.0356188780053428, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.490973472595215, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8817789554595947, "num_tokens": 310512956.0, "step": 8141 }, { "epoch": 1.0357460882839333, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.512630462646484, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8647499680519104, "num_tokens": 310547605.0, "step": 8142 }, { "epoch": 1.0358732985625239, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.551551818847656, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8639709949493408, "num_tokens": 310586991.0, "step": 8143 }, { "epoch": 1.0360005088411144, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.509293556213379, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8649320602416992, "num_tokens": 310621271.0, "step": 8144 }, { "epoch": 1.036127719119705, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.46607780456543, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8689119815826416, "num_tokens": 310662154.0, "step": 8145 }, { "epoch": 1.0362549293982954, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.50474739074707, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8764433264732361, "num_tokens": 310701924.0, "step": 8146 }, { "epoch": 1.036382139676886, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.503996849060059, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8714567422866821, "num_tokens": 310739688.0, "step": 8147 }, { "epoch": 1.0365093499554765, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.5153226852417, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8599717020988464, "num_tokens": 310773317.0, "step": 8148 }, { "epoch": 1.0366365602340668, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.527013778686523, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8642948865890503, "num_tokens": 310810854.0, "step": 8149 }, { "epoch": 1.0367637705126573, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.506936073303223, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8595821857452393, "num_tokens": 310846080.0, "step": 8150 }, { "epoch": 1.0368909807912479, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.534388542175293, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8431441187858582, "num_tokens": 310885032.0, "step": 8151 }, { "epoch": 1.0370181910698384, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.536558151245117, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8719271421432495, "num_tokens": 310924518.0, "step": 8152 }, { "epoch": 1.037145401348429, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.509145736694336, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8680096864700317, "num_tokens": 310961734.0, "step": 8153 }, { "epoch": 1.0372726116270194, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.486412048339844, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8688284158706665, "num_tokens": 311004541.0, "step": 8154 }, { "epoch": 1.03739982190561, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.508463859558105, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8747418522834778, "num_tokens": 311043947.0, "step": 8155 }, { "epoch": 1.0375270321842005, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.487900733947754, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8623353242874146, "num_tokens": 311083154.0, "step": 8156 }, { "epoch": 1.037654242462791, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.497703552246094, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8768724799156189, "num_tokens": 311121429.0, "step": 8157 }, { "epoch": 1.0377814527413816, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.544182777404785, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8506395220756531, "num_tokens": 311166578.0, "step": 8158 }, { "epoch": 1.037908663019972, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.504745483398438, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8688019514083862, "num_tokens": 311209645.0, "step": 8159 }, { "epoch": 1.0380358732985626, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.56474781036377, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8702456951141357, "num_tokens": 311248428.0, "step": 8160 }, { "epoch": 1.0381630835771531, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.436517715454102, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8672795295715332, "num_tokens": 311282040.0, "step": 8161 }, { "epoch": 1.0382902938557435, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.56068229675293, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8729767799377441, "num_tokens": 311320233.0, "step": 8162 }, { "epoch": 1.038417504134334, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.50300407409668, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8516190052032471, "num_tokens": 311357586.0, "step": 8163 }, { "epoch": 1.0385447144129245, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.509241104125977, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8652969598770142, "num_tokens": 311395545.0, "step": 8164 }, { "epoch": 1.038671924691515, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.457228660583496, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8652828931808472, "num_tokens": 311428785.0, "step": 8165 }, { "epoch": 1.0387991349701056, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.467937469482422, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8546361923217773, "num_tokens": 311469546.0, "step": 8166 }, { "epoch": 1.038926345248696, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.528892517089844, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8588584661483765, "num_tokens": 311504064.0, "step": 8167 }, { "epoch": 1.0390535555272866, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.564703941345215, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8479660153388977, "num_tokens": 311536600.0, "step": 8168 }, { "epoch": 1.0391807658058771, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.452436447143555, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8820960521697998, "num_tokens": 311573638.0, "step": 8169 }, { "epoch": 1.0393079760844677, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.558448791503906, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8585299253463745, "num_tokens": 311611085.0, "step": 8170 }, { "epoch": 1.0394351863630582, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.514471054077148, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8588780760765076, "num_tokens": 311655497.0, "step": 8171 }, { "epoch": 1.0395623966416487, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.528314590454102, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8507470488548279, "num_tokens": 311688167.0, "step": 8172 }, { "epoch": 1.0396896069202393, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.484317779541016, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8656590580940247, "num_tokens": 311724926.0, "step": 8173 }, { "epoch": 1.0398168171988296, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.52231502532959, "learning_rate": 1e-06, "loss": 0.4846, "mean_token_accuracy": 0.8467808365821838, "num_tokens": 311760729.0, "step": 8174 }, { "epoch": 1.03994402747742, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.50780200958252, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8658357858657837, "num_tokens": 311802138.0, "step": 8175 }, { "epoch": 1.0400712377560106, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.511833190917969, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8725652694702148, "num_tokens": 311838510.0, "step": 8176 }, { "epoch": 1.0401984480346012, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.583019256591797, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8585917949676514, "num_tokens": 311870875.0, "step": 8177 }, { "epoch": 1.0403256583131917, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.474014282226562, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8776606321334839, "num_tokens": 311905854.0, "step": 8178 }, { "epoch": 1.0404528685917822, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.53355884552002, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8628812432289124, "num_tokens": 311936381.0, "step": 8179 }, { "epoch": 1.0405800788703727, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.521029472351074, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.859927773475647, "num_tokens": 311972529.0, "step": 8180 }, { "epoch": 1.0407072891489633, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.56419849395752, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8842440843582153, "num_tokens": 312007883.0, "step": 8181 }, { "epoch": 1.0408344994275538, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.465517044067383, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.86198890209198, "num_tokens": 312041000.0, "step": 8182 }, { "epoch": 1.0409617097061443, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.491327285766602, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8691412210464478, "num_tokens": 312076011.0, "step": 8183 }, { "epoch": 1.0410889199847349, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.474777221679688, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.859271764755249, "num_tokens": 312115614.0, "step": 8184 }, { "epoch": 1.0412161302633254, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.524073600769043, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8594311475753784, "num_tokens": 312154111.0, "step": 8185 }, { "epoch": 1.0413433405419157, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.458868026733398, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8603787422180176, "num_tokens": 312191571.0, "step": 8186 }, { "epoch": 1.0414705508205062, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.458762168884277, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8585819005966187, "num_tokens": 312229205.0, "step": 8187 }, { "epoch": 1.0415977610990967, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.52367115020752, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8808106184005737, "num_tokens": 312266604.0, "step": 8188 }, { "epoch": 1.0417249713776873, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.470020294189453, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8559574484825134, "num_tokens": 312305222.0, "step": 8189 }, { "epoch": 1.0418521816562778, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.507913589477539, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8445504307746887, "num_tokens": 312345823.0, "step": 8190 }, { "epoch": 1.0419793919348683, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.531214714050293, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8564598560333252, "num_tokens": 312383569.0, "step": 8191 }, { "epoch": 1.0421066022134589, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.456186294555664, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8684584498405457, "num_tokens": 312423068.0, "step": 8192 }, { "epoch": 1.0422338124920494, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.507814407348633, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8599436283111572, "num_tokens": 312460569.0, "step": 8193 }, { "epoch": 1.04236102277064, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.5618896484375, "learning_rate": 1e-06, "loss": 0.4873, "mean_token_accuracy": 0.8439249992370605, "num_tokens": 312501654.0, "step": 8194 }, { "epoch": 1.0424882330492304, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.503335952758789, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8567622900009155, "num_tokens": 312543630.0, "step": 8195 }, { "epoch": 1.042615443327821, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.474949836730957, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8624902963638306, "num_tokens": 312579428.0, "step": 8196 }, { "epoch": 1.0427426536064115, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.508090019226074, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8798407316207886, "num_tokens": 312621388.0, "step": 8197 }, { "epoch": 1.0428698638850018, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.490987777709961, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8654792904853821, "num_tokens": 312665303.0, "step": 8198 }, { "epoch": 1.0429970741635923, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.505338668823242, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8545600771903992, "num_tokens": 312702354.0, "step": 8199 }, { "epoch": 1.0431242844421829, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.51755428314209, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8691871166229248, "num_tokens": 312737223.0, "step": 8200 }, { "epoch": 1.0432514947207734, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.492473602294922, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8658453226089478, "num_tokens": 312773353.0, "step": 8201 }, { "epoch": 1.043378704999364, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.543696403503418, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8466751575469971, "num_tokens": 312809230.0, "step": 8202 }, { "epoch": 1.0435059152779544, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.559144020080566, "learning_rate": 1e-06, "loss": 0.3558, "mean_token_accuracy": 0.8821502923965454, "num_tokens": 312841278.0, "step": 8203 }, { "epoch": 1.043633125556545, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.4350004196167, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8695853352546692, "num_tokens": 312888469.0, "step": 8204 }, { "epoch": 1.0437603358351355, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.542341232299805, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8721956014633179, "num_tokens": 312925446.0, "step": 8205 }, { "epoch": 1.043887546113726, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.476666450500488, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8639556169509888, "num_tokens": 312965425.0, "step": 8206 }, { "epoch": 1.0440147563923166, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.53976821899414, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8594587445259094, "num_tokens": 313000637.0, "step": 8207 }, { "epoch": 1.044141966670907, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.51054859161377, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8660295009613037, "num_tokens": 313035986.0, "step": 8208 }, { "epoch": 1.0442691769494976, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.487126350402832, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8544922471046448, "num_tokens": 313077027.0, "step": 8209 }, { "epoch": 1.0443963872280881, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.458647727966309, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8754329681396484, "num_tokens": 313121733.0, "step": 8210 }, { "epoch": 1.0445235975066784, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.532779693603516, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8601005673408508, "num_tokens": 313165542.0, "step": 8211 }, { "epoch": 1.044650807785269, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.483817100524902, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8539441823959351, "num_tokens": 313201082.0, "step": 8212 }, { "epoch": 1.0447780180638595, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.52643871307373, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8763698935508728, "num_tokens": 313241765.0, "step": 8213 }, { "epoch": 1.04490522834245, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.550783157348633, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8628413677215576, "num_tokens": 313282471.0, "step": 8214 }, { "epoch": 1.0450324386210406, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.527240753173828, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8604605793952942, "num_tokens": 313331505.0, "step": 8215 }, { "epoch": 1.045159648899631, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.519237518310547, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8658326864242554, "num_tokens": 313369823.0, "step": 8216 }, { "epoch": 1.0452868591782216, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.539833068847656, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8658616542816162, "num_tokens": 313404255.0, "step": 8217 }, { "epoch": 1.0454140694568121, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.562081336975098, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.876224160194397, "num_tokens": 313439368.0, "step": 8218 }, { "epoch": 1.0455412797354027, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.515567779541016, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8569921255111694, "num_tokens": 313479475.0, "step": 8219 }, { "epoch": 1.0456684900139932, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.573209762573242, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8554559350013733, "num_tokens": 313514604.0, "step": 8220 }, { "epoch": 1.0457957002925837, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.483026504516602, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8614369630813599, "num_tokens": 313548907.0, "step": 8221 }, { "epoch": 1.0459229105711743, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.523504257202148, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8641132712364197, "num_tokens": 313586615.0, "step": 8222 }, { "epoch": 1.0460501208497646, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.503233909606934, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8558582663536072, "num_tokens": 313625202.0, "step": 8223 }, { "epoch": 1.046177331128355, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.540060043334961, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.857875406742096, "num_tokens": 313663257.0, "step": 8224 }, { "epoch": 1.0463045414069456, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.494166374206543, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8690115213394165, "num_tokens": 313701430.0, "step": 8225 }, { "epoch": 1.0464317516855361, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.563289642333984, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8764898180961609, "num_tokens": 313741286.0, "step": 8226 }, { "epoch": 1.0465589619641267, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.561779022216797, "learning_rate": 1e-06, "loss": 0.4888, "mean_token_accuracy": 0.8496404886245728, "num_tokens": 313779115.0, "step": 8227 }, { "epoch": 1.0466861722427172, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.476724624633789, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8696614503860474, "num_tokens": 313822331.0, "step": 8228 }, { "epoch": 1.0468133825213077, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.493636131286621, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8592746257781982, "num_tokens": 313863545.0, "step": 8229 }, { "epoch": 1.0469405927998983, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.554106712341309, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8845293521881104, "num_tokens": 313894956.0, "step": 8230 }, { "epoch": 1.0470678030784888, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.592536926269531, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8731478452682495, "num_tokens": 313927920.0, "step": 8231 }, { "epoch": 1.0471950133570793, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.47564697265625, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8737325668334961, "num_tokens": 313963230.0, "step": 8232 }, { "epoch": 1.0473222236356698, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.592964172363281, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8571571111679077, "num_tokens": 313995951.0, "step": 8233 }, { "epoch": 1.0474494339142604, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.489099502563477, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8653827905654907, "num_tokens": 314040928.0, "step": 8234 }, { "epoch": 1.0475766441928507, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.50525951385498, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8576587438583374, "num_tokens": 314080329.0, "step": 8235 }, { "epoch": 1.0477038544714412, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.583191871643066, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8670178651809692, "num_tokens": 314117107.0, "step": 8236 }, { "epoch": 1.0478310647500317, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.52470874786377, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8632164001464844, "num_tokens": 314148077.0, "step": 8237 }, { "epoch": 1.0479582750286223, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.499353408813477, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.856259822845459, "num_tokens": 314182064.0, "step": 8238 }, { "epoch": 1.0480854853072128, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.500473976135254, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8571416139602661, "num_tokens": 314221667.0, "step": 8239 }, { "epoch": 1.0482126955858033, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.500777244567871, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8615764379501343, "num_tokens": 314255225.0, "step": 8240 }, { "epoch": 1.0483399058643939, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.480925559997559, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8653439879417419, "num_tokens": 314288011.0, "step": 8241 }, { "epoch": 1.0484671161429844, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.479612350463867, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8621490001678467, "num_tokens": 314332461.0, "step": 8242 }, { "epoch": 1.048594326421575, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.561912536621094, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8608285188674927, "num_tokens": 314367967.0, "step": 8243 }, { "epoch": 1.0487215367001654, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.480175971984863, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8745260238647461, "num_tokens": 314404487.0, "step": 8244 }, { "epoch": 1.048848746978756, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.51850700378418, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8621096611022949, "num_tokens": 314444731.0, "step": 8245 }, { "epoch": 1.0489759572573465, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.472009658813477, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8652960658073425, "num_tokens": 314479720.0, "step": 8246 }, { "epoch": 1.0491031675359368, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.49375057220459, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8822052478790283, "num_tokens": 314515613.0, "step": 8247 }, { "epoch": 1.0492303778145273, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.545969009399414, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8637782335281372, "num_tokens": 314556519.0, "step": 8248 }, { "epoch": 1.0493575880931179, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.451327323913574, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.875600278377533, "num_tokens": 314590803.0, "step": 8249 }, { "epoch": 1.0494847983717084, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.525654792785645, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8662462830543518, "num_tokens": 314631347.0, "step": 8250 }, { "epoch": 1.049612008650299, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.481030464172363, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8647522330284119, "num_tokens": 314673739.0, "step": 8251 }, { "epoch": 1.0497392189288894, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.543512344360352, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8653262853622437, "num_tokens": 314718498.0, "step": 8252 }, { "epoch": 1.04986642920748, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.539453506469727, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8700482845306396, "num_tokens": 314758755.0, "step": 8253 }, { "epoch": 1.0499936394860705, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.57907485961914, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8685968518257141, "num_tokens": 314793644.0, "step": 8254 }, { "epoch": 1.050120849764661, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.496793746948242, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8617057800292969, "num_tokens": 314828045.0, "step": 8255 }, { "epoch": 1.0502480600432516, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.527262687683105, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.868350625038147, "num_tokens": 314866088.0, "step": 8256 }, { "epoch": 1.050375270321842, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.546173095703125, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8564801216125488, "num_tokens": 314904757.0, "step": 8257 }, { "epoch": 1.0505024806004326, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.536096572875977, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8575950860977173, "num_tokens": 314941410.0, "step": 8258 }, { "epoch": 1.0506296908790231, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.523065567016602, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.874054491519928, "num_tokens": 314982060.0, "step": 8259 }, { "epoch": 1.0507569011576134, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.567522048950195, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8558589220046997, "num_tokens": 315021812.0, "step": 8260 }, { "epoch": 1.050884111436204, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.531900405883789, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.861062228679657, "num_tokens": 315059520.0, "step": 8261 }, { "epoch": 1.0510113217147945, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.505173683166504, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.860687792301178, "num_tokens": 315098392.0, "step": 8262 }, { "epoch": 1.051138531993385, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.490457534790039, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8543514013290405, "num_tokens": 315137375.0, "step": 8263 }, { "epoch": 1.0512657422719756, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.5664701461792, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8694283962249756, "num_tokens": 315171308.0, "step": 8264 }, { "epoch": 1.051392952550566, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.47675895690918, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8659976124763489, "num_tokens": 315212176.0, "step": 8265 }, { "epoch": 1.0515201628291566, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.523327827453613, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8531812429428101, "num_tokens": 315251321.0, "step": 8266 }, { "epoch": 1.0516473731077471, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.592092514038086, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.851105272769928, "num_tokens": 315288780.0, "step": 8267 }, { "epoch": 1.0517745833863377, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.495594024658203, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.8870682716369629, "num_tokens": 315325630.0, "step": 8268 }, { "epoch": 1.0519017936649282, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.558323860168457, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8593102097511292, "num_tokens": 315362887.0, "step": 8269 }, { "epoch": 1.0520290039435187, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.581629753112793, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8774697780609131, "num_tokens": 315397552.0, "step": 8270 }, { "epoch": 1.0521562142221093, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.57376480102539, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8750841617584229, "num_tokens": 315436785.0, "step": 8271 }, { "epoch": 1.0522834245006996, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.546496391296387, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8750898241996765, "num_tokens": 315467849.0, "step": 8272 }, { "epoch": 1.05241063477929, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.503462791442871, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8726913332939148, "num_tokens": 315501496.0, "step": 8273 }, { "epoch": 1.0525378450578806, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.539007186889648, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8731290102005005, "num_tokens": 315540878.0, "step": 8274 }, { "epoch": 1.0526650553364711, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.498863220214844, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8800872564315796, "num_tokens": 315578139.0, "step": 8275 }, { "epoch": 1.0527922656150617, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.52907657623291, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8662147521972656, "num_tokens": 315615845.0, "step": 8276 }, { "epoch": 1.0529194758936522, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.551521301269531, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8575292825698853, "num_tokens": 315645407.0, "step": 8277 }, { "epoch": 1.0530466861722427, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.55545711517334, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8728234767913818, "num_tokens": 315687922.0, "step": 8278 }, { "epoch": 1.0531738964508333, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.5487642288208, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8666293621063232, "num_tokens": 315723730.0, "step": 8279 }, { "epoch": 1.0533011067294238, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.571266174316406, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8772954940795898, "num_tokens": 315765662.0, "step": 8280 }, { "epoch": 1.0534283170080143, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.51147174835205, "learning_rate": 1e-06, "loss": 0.489, "mean_token_accuracy": 0.8404390811920166, "num_tokens": 315808668.0, "step": 8281 }, { "epoch": 1.0535555272866048, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.585673332214355, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.861137330532074, "num_tokens": 315844878.0, "step": 8282 }, { "epoch": 1.0536827375651954, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.592920303344727, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8781882524490356, "num_tokens": 315879472.0, "step": 8283 }, { "epoch": 1.0538099478437857, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.510586738586426, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8603875637054443, "num_tokens": 315915496.0, "step": 8284 }, { "epoch": 1.0539371581223762, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.513495445251465, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8678480386734009, "num_tokens": 315954317.0, "step": 8285 }, { "epoch": 1.0540643684009667, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.449448585510254, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8616878390312195, "num_tokens": 315995518.0, "step": 8286 }, { "epoch": 1.0541915786795573, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.533702850341797, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8790018558502197, "num_tokens": 316040475.0, "step": 8287 }, { "epoch": 1.0543187889581478, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.504244804382324, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8640512228012085, "num_tokens": 316083003.0, "step": 8288 }, { "epoch": 1.0544459992367383, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.517598152160645, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8654890060424805, "num_tokens": 316124687.0, "step": 8289 }, { "epoch": 1.0545732095153288, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.49113941192627, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8533009886741638, "num_tokens": 316170052.0, "step": 8290 }, { "epoch": 1.0547004197939194, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.537712097167969, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8522312641143799, "num_tokens": 316211941.0, "step": 8291 }, { "epoch": 1.05482763007251, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.490912437438965, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8693810701370239, "num_tokens": 316252592.0, "step": 8292 }, { "epoch": 1.0549548403511004, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.52413272857666, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8734673261642456, "num_tokens": 316287250.0, "step": 8293 }, { "epoch": 1.055082050629691, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.47915267944336, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8690282106399536, "num_tokens": 316326308.0, "step": 8294 }, { "epoch": 1.0552092609082815, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.50096607208252, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.870031476020813, "num_tokens": 316364519.0, "step": 8295 }, { "epoch": 1.0553364711868718, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.497221946716309, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8707472085952759, "num_tokens": 316401165.0, "step": 8296 }, { "epoch": 1.0554636814654623, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.530808448791504, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8604429960250854, "num_tokens": 316440245.0, "step": 8297 }, { "epoch": 1.0555908917440529, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.499105453491211, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8737188577651978, "num_tokens": 316484969.0, "step": 8298 }, { "epoch": 1.0557181020226434, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.55286979675293, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8837668299674988, "num_tokens": 316520510.0, "step": 8299 }, { "epoch": 1.055845312301234, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.453754425048828, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8499940037727356, "num_tokens": 316561946.0, "step": 8300 }, { "epoch": 1.0559725225798244, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.539520263671875, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8833513259887695, "num_tokens": 316605489.0, "step": 8301 }, { "epoch": 1.056099732858415, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.440873146057129, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.848540186882019, "num_tokens": 316648590.0, "step": 8302 }, { "epoch": 1.0562269431370055, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.526016235351562, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.878939151763916, "num_tokens": 316688169.0, "step": 8303 }, { "epoch": 1.056354153415596, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.500833511352539, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.8460701704025269, "num_tokens": 316724031.0, "step": 8304 }, { "epoch": 1.0564813636941865, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.514395713806152, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8656858205795288, "num_tokens": 316769762.0, "step": 8305 }, { "epoch": 1.056608573972777, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.489910125732422, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8736773729324341, "num_tokens": 316814156.0, "step": 8306 }, { "epoch": 1.0567357842513676, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.551066398620605, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8746504187583923, "num_tokens": 316852784.0, "step": 8307 }, { "epoch": 1.0568629945299581, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.539544105529785, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8589966297149658, "num_tokens": 316891992.0, "step": 8308 }, { "epoch": 1.0569902048085484, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.533935546875, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8805019855499268, "num_tokens": 316932246.0, "step": 8309 }, { "epoch": 1.057117415087139, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.52103328704834, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8575167655944824, "num_tokens": 316972017.0, "step": 8310 }, { "epoch": 1.0572446253657295, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.538653373718262, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8610173463821411, "num_tokens": 317008631.0, "step": 8311 }, { "epoch": 1.05737183564432, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.490235328674316, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8709318041801453, "num_tokens": 317048616.0, "step": 8312 }, { "epoch": 1.0574990459229106, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.587139129638672, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8592263460159302, "num_tokens": 317088675.0, "step": 8313 }, { "epoch": 1.057626256201501, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.521443367004395, "learning_rate": 1e-06, "loss": 0.347, "mean_token_accuracy": 0.8863648176193237, "num_tokens": 317120351.0, "step": 8314 }, { "epoch": 1.0577534664800916, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.496700286865234, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.859503448009491, "num_tokens": 317159027.0, "step": 8315 }, { "epoch": 1.0578806767586821, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.591348648071289, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8663761615753174, "num_tokens": 317188892.0, "step": 8316 }, { "epoch": 1.0580078870372727, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.518997192382812, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8709001541137695, "num_tokens": 317229026.0, "step": 8317 }, { "epoch": 1.0581350973158632, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.520413398742676, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8645023107528687, "num_tokens": 317272946.0, "step": 8318 }, { "epoch": 1.0582623075944537, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.594987869262695, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8655978441238403, "num_tokens": 317314625.0, "step": 8319 }, { "epoch": 1.058389517873044, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.530501365661621, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8756277561187744, "num_tokens": 317351450.0, "step": 8320 }, { "epoch": 1.0585167281516346, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.620348930358887, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8500654101371765, "num_tokens": 317391519.0, "step": 8321 }, { "epoch": 1.058643938430225, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.508978843688965, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.859757661819458, "num_tokens": 317430672.0, "step": 8322 }, { "epoch": 1.0587711487088156, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.540270805358887, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8676809072494507, "num_tokens": 317470203.0, "step": 8323 }, { "epoch": 1.0588983589874061, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.564827919006348, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8684777617454529, "num_tokens": 317509594.0, "step": 8324 }, { "epoch": 1.0590255692659967, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.536518096923828, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.863489031791687, "num_tokens": 317549735.0, "step": 8325 }, { "epoch": 1.0591527795445872, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.604222297668457, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8643561601638794, "num_tokens": 317585064.0, "step": 8326 }, { "epoch": 1.0592799898231777, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.487468719482422, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8810077905654907, "num_tokens": 317622289.0, "step": 8327 }, { "epoch": 1.0594072001017683, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.565009117126465, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8546502590179443, "num_tokens": 317662872.0, "step": 8328 }, { "epoch": 1.0595344103803588, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.58246898651123, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8663722276687622, "num_tokens": 317701418.0, "step": 8329 }, { "epoch": 1.0596616206589493, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.487737655639648, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8786780834197998, "num_tokens": 317741416.0, "step": 8330 }, { "epoch": 1.0597888309375398, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.55435848236084, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8705500364303589, "num_tokens": 317778623.0, "step": 8331 }, { "epoch": 1.0599160412161304, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.540807723999023, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.86782306432724, "num_tokens": 317816577.0, "step": 8332 }, { "epoch": 1.0600432514947207, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.542007446289062, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8644609451293945, "num_tokens": 317851532.0, "step": 8333 }, { "epoch": 1.0601704617733112, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.544610023498535, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8664190173149109, "num_tokens": 317886789.0, "step": 8334 }, { "epoch": 1.0602976720519017, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.493932723999023, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8680002689361572, "num_tokens": 317929511.0, "step": 8335 }, { "epoch": 1.0604248823304923, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.578008651733398, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.864425539970398, "num_tokens": 317963767.0, "step": 8336 }, { "epoch": 1.0605520926090828, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.51588249206543, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8832621574401855, "num_tokens": 318000458.0, "step": 8337 }, { "epoch": 1.0606793028876733, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.52044677734375, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8836535215377808, "num_tokens": 318040394.0, "step": 8338 }, { "epoch": 1.0608065131662638, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.615896224975586, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.86216139793396, "num_tokens": 318080859.0, "step": 8339 }, { "epoch": 1.0609337234448544, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.49925708770752, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8827965259552002, "num_tokens": 318115378.0, "step": 8340 }, { "epoch": 1.061060933723445, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.611403465270996, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8582720756530762, "num_tokens": 318150351.0, "step": 8341 }, { "epoch": 1.0611881440020354, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.527946472167969, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8660576343536377, "num_tokens": 318192427.0, "step": 8342 }, { "epoch": 1.061315354280626, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.560613632202148, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8534387946128845, "num_tokens": 318232150.0, "step": 8343 }, { "epoch": 1.0614425645592165, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.542553901672363, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8626915216445923, "num_tokens": 318267742.0, "step": 8344 }, { "epoch": 1.0615697748378068, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.531607627868652, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8706450462341309, "num_tokens": 318307974.0, "step": 8345 }, { "epoch": 1.0616969851163973, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.581071853637695, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8483466506004333, "num_tokens": 318349880.0, "step": 8346 }, { "epoch": 1.0618241953949878, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.58106517791748, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8647043704986572, "num_tokens": 318383869.0, "step": 8347 }, { "epoch": 1.0619514056735784, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.595733642578125, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8820458054542542, "num_tokens": 318419659.0, "step": 8348 }, { "epoch": 1.062078615952169, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.483108520507812, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8749514222145081, "num_tokens": 318460795.0, "step": 8349 }, { "epoch": 1.0622058262307594, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.671375274658203, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8567830324172974, "num_tokens": 318499451.0, "step": 8350 }, { "epoch": 1.06233303650935, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.514957427978516, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8695546388626099, "num_tokens": 318533413.0, "step": 8351 }, { "epoch": 1.0624602467879405, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.573431968688965, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8552303910255432, "num_tokens": 318572533.0, "step": 8352 }, { "epoch": 1.062587457066531, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.55907917022705, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8680728673934937, "num_tokens": 318609442.0, "step": 8353 }, { "epoch": 1.0627146673451215, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.60973834991455, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8809204697608948, "num_tokens": 318643475.0, "step": 8354 }, { "epoch": 1.062841877623712, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.560359954833984, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8699827790260315, "num_tokens": 318684223.0, "step": 8355 }, { "epoch": 1.0629690879023026, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.591320991516113, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8669173121452332, "num_tokens": 318717148.0, "step": 8356 }, { "epoch": 1.0630962981808931, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.640990257263184, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8773581981658936, "num_tokens": 318758929.0, "step": 8357 }, { "epoch": 1.0632235084594834, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.558953285217285, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8637639284133911, "num_tokens": 318796438.0, "step": 8358 }, { "epoch": 1.063350718738074, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.60959243774414, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8538094758987427, "num_tokens": 318835732.0, "step": 8359 }, { "epoch": 1.0634779290166645, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.582856178283691, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8650475740432739, "num_tokens": 318872151.0, "step": 8360 }, { "epoch": 1.063605139295255, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.645896911621094, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8556864261627197, "num_tokens": 318911956.0, "step": 8361 }, { "epoch": 1.0637323495738455, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.522045135498047, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8675537109375, "num_tokens": 318945593.0, "step": 8362 }, { "epoch": 1.063859559852436, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.658754348754883, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8761316537857056, "num_tokens": 318976670.0, "step": 8363 }, { "epoch": 1.0639867701310266, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.504295349121094, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8676886558532715, "num_tokens": 319007798.0, "step": 8364 }, { "epoch": 1.0641139804096171, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.602703094482422, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8554248213768005, "num_tokens": 319044808.0, "step": 8365 }, { "epoch": 1.0642411906882077, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.574007034301758, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8855578303337097, "num_tokens": 319083387.0, "step": 8366 }, { "epoch": 1.0643684009667982, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.455500602722168, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8818103671073914, "num_tokens": 319121897.0, "step": 8367 }, { "epoch": 1.0644956112453887, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.61838436126709, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.854787290096283, "num_tokens": 319157190.0, "step": 8368 }, { "epoch": 1.064622821523979, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.568634033203125, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8771275281906128, "num_tokens": 319188353.0, "step": 8369 }, { "epoch": 1.0647500318025696, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.513177871704102, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8681293725967407, "num_tokens": 319225536.0, "step": 8370 }, { "epoch": 1.06487724208116, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.628405570983887, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8734409809112549, "num_tokens": 319265114.0, "step": 8371 }, { "epoch": 1.0650044523597506, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.520055770874023, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.877118706703186, "num_tokens": 319309280.0, "step": 8372 }, { "epoch": 1.0651316626383411, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.575150489807129, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.877774178981781, "num_tokens": 319346114.0, "step": 8373 }, { "epoch": 1.0652588729169317, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.564173698425293, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.873542845249176, "num_tokens": 319389001.0, "step": 8374 }, { "epoch": 1.0653860831955222, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.549263954162598, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.873928964138031, "num_tokens": 319431464.0, "step": 8375 }, { "epoch": 1.0655132934741127, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.560335159301758, "learning_rate": 1e-06, "loss": 0.4889, "mean_token_accuracy": 0.8432654142379761, "num_tokens": 319475638.0, "step": 8376 }, { "epoch": 1.0656405037527032, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.488082885742188, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8765044212341309, "num_tokens": 319515983.0, "step": 8377 }, { "epoch": 1.0657677140312938, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.644606590270996, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8636140823364258, "num_tokens": 319552152.0, "step": 8378 }, { "epoch": 1.0658949243098843, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.483009338378906, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8583589792251587, "num_tokens": 319592963.0, "step": 8379 }, { "epoch": 1.0660221345884748, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.592939376831055, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8531017899513245, "num_tokens": 319635136.0, "step": 8380 }, { "epoch": 1.0661493448670654, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.576549530029297, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8691544532775879, "num_tokens": 319670118.0, "step": 8381 }, { "epoch": 1.0662765551456557, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.540288925170898, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8735864162445068, "num_tokens": 319711057.0, "step": 8382 }, { "epoch": 1.0664037654242462, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.582836151123047, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8655027151107788, "num_tokens": 319746451.0, "step": 8383 }, { "epoch": 1.0665309757028367, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.517352104187012, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8605847954750061, "num_tokens": 319792149.0, "step": 8384 }, { "epoch": 1.0666581859814273, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.50994873046875, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8578500747680664, "num_tokens": 319826508.0, "step": 8385 }, { "epoch": 1.0667853962600178, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.53752613067627, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.868418276309967, "num_tokens": 319867000.0, "step": 8386 }, { "epoch": 1.0669126065386083, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.70798110961914, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8488147854804993, "num_tokens": 319904476.0, "step": 8387 }, { "epoch": 1.0670398168171988, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.617344856262207, "learning_rate": 1e-06, "loss": 0.4882, "mean_token_accuracy": 0.8429679870605469, "num_tokens": 319943279.0, "step": 8388 }, { "epoch": 1.0671670270957894, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.506455421447754, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8631894588470459, "num_tokens": 319986610.0, "step": 8389 }, { "epoch": 1.06729423737438, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.623011589050293, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.860227108001709, "num_tokens": 320022702.0, "step": 8390 }, { "epoch": 1.0674214476529704, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.505912780761719, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8688815832138062, "num_tokens": 320067918.0, "step": 8391 }, { "epoch": 1.067548657931561, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.582362174987793, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.857946515083313, "num_tokens": 320110455.0, "step": 8392 }, { "epoch": 1.0676758682101515, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.525429725646973, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8828124403953552, "num_tokens": 320147675.0, "step": 8393 }, { "epoch": 1.0678030784887418, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.582963943481445, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8529853224754333, "num_tokens": 320183882.0, "step": 8394 }, { "epoch": 1.0679302887673323, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.570277214050293, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8555009365081787, "num_tokens": 320222215.0, "step": 8395 }, { "epoch": 1.0680574990459228, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.550058364868164, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8592679500579834, "num_tokens": 320262418.0, "step": 8396 }, { "epoch": 1.0681847093245134, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.657495498657227, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8593786954879761, "num_tokens": 320297304.0, "step": 8397 }, { "epoch": 1.068311919603104, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.546992301940918, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8688985109329224, "num_tokens": 320330919.0, "step": 8398 }, { "epoch": 1.0684391298816944, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.533920288085938, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8657349348068237, "num_tokens": 320367602.0, "step": 8399 }, { "epoch": 1.068566340160285, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.521252632141113, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8661878108978271, "num_tokens": 320403875.0, "step": 8400 }, { "epoch": 1.0686935504388755, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.568075180053711, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8700906038284302, "num_tokens": 320439149.0, "step": 8401 }, { "epoch": 1.068820760717466, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.557304382324219, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.877860963344574, "num_tokens": 320476786.0, "step": 8402 }, { "epoch": 1.0689479709960565, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.52009391784668, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8505086302757263, "num_tokens": 320516945.0, "step": 8403 }, { "epoch": 1.069075181274647, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.530969619750977, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8811266422271729, "num_tokens": 320552801.0, "step": 8404 }, { "epoch": 1.0692023915532376, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.55698299407959, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8692711591720581, "num_tokens": 320593605.0, "step": 8405 }, { "epoch": 1.0693296018318281, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.615503311157227, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8732470273971558, "num_tokens": 320624476.0, "step": 8406 }, { "epoch": 1.0694568121104184, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.517683982849121, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8736087679862976, "num_tokens": 320667000.0, "step": 8407 }, { "epoch": 1.069584022389009, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.577193260192871, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8704396486282349, "num_tokens": 320699286.0, "step": 8408 }, { "epoch": 1.0697112326675995, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.614812850952148, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8728915452957153, "num_tokens": 320732520.0, "step": 8409 }, { "epoch": 1.06983844294619, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.613210678100586, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.855427622795105, "num_tokens": 320773425.0, "step": 8410 }, { "epoch": 1.0699656532247805, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.547847747802734, "learning_rate": 1e-06, "loss": 0.4963, "mean_token_accuracy": 0.8444163799285889, "num_tokens": 320810859.0, "step": 8411 }, { "epoch": 1.070092863503371, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.647180557250977, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8697608709335327, "num_tokens": 320850266.0, "step": 8412 }, { "epoch": 1.0702200737819616, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.542933464050293, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8648070096969604, "num_tokens": 320892812.0, "step": 8413 }, { "epoch": 1.0703472840605521, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.640063285827637, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8780308961868286, "num_tokens": 320929563.0, "step": 8414 }, { "epoch": 1.0704744943391427, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.571229934692383, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8736223578453064, "num_tokens": 320964010.0, "step": 8415 }, { "epoch": 1.0706017046177332, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.52940559387207, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8637726902961731, "num_tokens": 321006405.0, "step": 8416 }, { "epoch": 1.0707289148963237, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.549256324768066, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8611405491828918, "num_tokens": 321050146.0, "step": 8417 }, { "epoch": 1.070856125174914, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.603535652160645, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8660598993301392, "num_tokens": 321086857.0, "step": 8418 }, { "epoch": 1.0709833354535045, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.534884452819824, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8721113204956055, "num_tokens": 321127310.0, "step": 8419 }, { "epoch": 1.071110545732095, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.577744483947754, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8568487763404846, "num_tokens": 321165881.0, "step": 8420 }, { "epoch": 1.0712377560106856, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.576321601867676, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8595764636993408, "num_tokens": 321204494.0, "step": 8421 }, { "epoch": 1.0713649662892761, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.592026710510254, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8571910858154297, "num_tokens": 321243639.0, "step": 8422 }, { "epoch": 1.0714921765678667, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.55781364440918, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8596867322921753, "num_tokens": 321281399.0, "step": 8423 }, { "epoch": 1.0716193868464572, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.632345199584961, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8529571294784546, "num_tokens": 321316539.0, "step": 8424 }, { "epoch": 1.0717465971250477, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.582132339477539, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8612547516822815, "num_tokens": 321348618.0, "step": 8425 }, { "epoch": 1.0718738074036382, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.586788177490234, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8724704384803772, "num_tokens": 321381674.0, "step": 8426 }, { "epoch": 1.0720010176822288, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.53134822845459, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8826595544815063, "num_tokens": 321420342.0, "step": 8427 }, { "epoch": 1.0721282279608193, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.567481994628906, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8630672693252563, "num_tokens": 321450770.0, "step": 8428 }, { "epoch": 1.0722554382394098, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.612622261047363, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8695623278617859, "num_tokens": 321491450.0, "step": 8429 }, { "epoch": 1.0723826485180004, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.552993774414062, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8599399924278259, "num_tokens": 321531774.0, "step": 8430 }, { "epoch": 1.0725098587965907, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.571836471557617, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8661690950393677, "num_tokens": 321567692.0, "step": 8431 }, { "epoch": 1.0726370690751812, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.625374794006348, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8613598346710205, "num_tokens": 321601168.0, "step": 8432 }, { "epoch": 1.0727642793537717, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.549905776977539, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.8477795720100403, "num_tokens": 321636100.0, "step": 8433 }, { "epoch": 1.0728914896323622, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.60543441772461, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8674901723861694, "num_tokens": 321674365.0, "step": 8434 }, { "epoch": 1.0730186999109528, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.552593231201172, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.877525269985199, "num_tokens": 321712267.0, "step": 8435 }, { "epoch": 1.0731459101895433, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.625855445861816, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8644869923591614, "num_tokens": 321749955.0, "step": 8436 }, { "epoch": 1.0732731204681338, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.531233787536621, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8739501237869263, "num_tokens": 321788522.0, "step": 8437 }, { "epoch": 1.0734003307467244, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.616251945495605, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8638627529144287, "num_tokens": 321826314.0, "step": 8438 }, { "epoch": 1.073527541025315, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.625432968139648, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.86210036277771, "num_tokens": 321864851.0, "step": 8439 }, { "epoch": 1.0736547513039054, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.626198768615723, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8754684329032898, "num_tokens": 321901001.0, "step": 8440 }, { "epoch": 1.073781961582496, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.598223686218262, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8647079467773438, "num_tokens": 321941368.0, "step": 8441 }, { "epoch": 1.0739091718610865, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.613269805908203, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8691806793212891, "num_tokens": 321978268.0, "step": 8442 }, { "epoch": 1.0740363821396768, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.578084945678711, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8727679252624512, "num_tokens": 322016969.0, "step": 8443 }, { "epoch": 1.0741635924182673, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.588260650634766, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8589992523193359, "num_tokens": 322059362.0, "step": 8444 }, { "epoch": 1.0742908026968578, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.615679740905762, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.846904456615448, "num_tokens": 322097334.0, "step": 8445 }, { "epoch": 1.0744180129754484, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.60934829711914, "learning_rate": 1e-06, "loss": 0.3406, "mean_token_accuracy": 0.8852832317352295, "num_tokens": 322121988.0, "step": 8446 }, { "epoch": 1.074545223254039, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.621954917907715, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8609800338745117, "num_tokens": 322156602.0, "step": 8447 }, { "epoch": 1.0746724335326294, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.564952850341797, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8621184825897217, "num_tokens": 322194475.0, "step": 8448 }, { "epoch": 1.07479964381122, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.595417976379395, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.873316764831543, "num_tokens": 322232158.0, "step": 8449 }, { "epoch": 1.0749268540898105, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.579134941101074, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8742712736129761, "num_tokens": 322272170.0, "step": 8450 }, { "epoch": 1.075054064368401, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.589714050292969, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8706591129302979, "num_tokens": 322305496.0, "step": 8451 }, { "epoch": 1.0751812746469915, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.597602844238281, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8705023527145386, "num_tokens": 322347552.0, "step": 8452 }, { "epoch": 1.075308484925582, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.599934577941895, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8691644072532654, "num_tokens": 322386471.0, "step": 8453 }, { "epoch": 1.0754356952041726, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.612144470214844, "learning_rate": 1e-06, "loss": 0.5187, "mean_token_accuracy": 0.8317720890045166, "num_tokens": 322431019.0, "step": 8454 }, { "epoch": 1.0755629054827631, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.569231986999512, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8844728469848633, "num_tokens": 322471525.0, "step": 8455 }, { "epoch": 1.0756901157613534, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.660511016845703, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.868347704410553, "num_tokens": 322516594.0, "step": 8456 }, { "epoch": 1.075817326039944, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.512250900268555, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8712139129638672, "num_tokens": 322557837.0, "step": 8457 }, { "epoch": 1.0759445363185345, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.562596321105957, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8621264696121216, "num_tokens": 322594165.0, "step": 8458 }, { "epoch": 1.076071746597125, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.698987007141113, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8485702872276306, "num_tokens": 322637083.0, "step": 8459 }, { "epoch": 1.0761989568757155, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.545329093933105, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8651921153068542, "num_tokens": 322673216.0, "step": 8460 }, { "epoch": 1.076326167154306, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.686527252197266, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8691961765289307, "num_tokens": 322709150.0, "step": 8461 }, { "epoch": 1.0764533774328966, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.527531623840332, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8728910684585571, "num_tokens": 322747893.0, "step": 8462 }, { "epoch": 1.0765805877114871, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.655078887939453, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8487992286682129, "num_tokens": 322781667.0, "step": 8463 }, { "epoch": 1.0767077979900777, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.638591766357422, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8655158877372742, "num_tokens": 322818008.0, "step": 8464 }, { "epoch": 1.0768350082686682, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.517702102661133, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8460332751274109, "num_tokens": 322856528.0, "step": 8465 }, { "epoch": 1.0769622185472587, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.659353256225586, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8692117929458618, "num_tokens": 322888798.0, "step": 8466 }, { "epoch": 1.077089428825849, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.602423667907715, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8679838180541992, "num_tokens": 322923301.0, "step": 8467 }, { "epoch": 1.0772166391044395, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.577326774597168, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8773782849311829, "num_tokens": 322964253.0, "step": 8468 }, { "epoch": 1.07734384938303, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.593283653259277, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8636949062347412, "num_tokens": 323006926.0, "step": 8469 }, { "epoch": 1.0774710596616206, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.561367988586426, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8785713315010071, "num_tokens": 323047541.0, "step": 8470 }, { "epoch": 1.0775982699402111, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.660449028015137, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8577998876571655, "num_tokens": 323084272.0, "step": 8471 }, { "epoch": 1.0777254802188017, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.681123733520508, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8773266077041626, "num_tokens": 323116698.0, "step": 8472 }, { "epoch": 1.0778526904973922, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.529748916625977, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8732234239578247, "num_tokens": 323158836.0, "step": 8473 }, { "epoch": 1.0779799007759827, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.634598731994629, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8616820573806763, "num_tokens": 323193777.0, "step": 8474 }, { "epoch": 1.0781071110545732, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.509675025939941, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8669009804725647, "num_tokens": 323232546.0, "step": 8475 }, { "epoch": 1.0782343213331638, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.631842613220215, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8661186695098877, "num_tokens": 323269403.0, "step": 8476 }, { "epoch": 1.0783615316117543, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.551766395568848, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8789390325546265, "num_tokens": 323305982.0, "step": 8477 }, { "epoch": 1.0784887418903448, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.597169876098633, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.86228346824646, "num_tokens": 323341032.0, "step": 8478 }, { "epoch": 1.0786159521689354, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.62418270111084, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8604793548583984, "num_tokens": 323381647.0, "step": 8479 }, { "epoch": 1.0787431624475257, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.593554496765137, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8699842691421509, "num_tokens": 323417185.0, "step": 8480 }, { "epoch": 1.0788703727261162, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.619767189025879, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8661187291145325, "num_tokens": 323454624.0, "step": 8481 }, { "epoch": 1.0789975830047067, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.564525604248047, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.866284191608429, "num_tokens": 323492862.0, "step": 8482 }, { "epoch": 1.0791247932832972, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.591192245483398, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8629515171051025, "num_tokens": 323536012.0, "step": 8483 }, { "epoch": 1.0792520035618878, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.58910846710205, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.870363712310791, "num_tokens": 323573495.0, "step": 8484 }, { "epoch": 1.0793792138404783, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.60513687133789, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8764491081237793, "num_tokens": 323610900.0, "step": 8485 }, { "epoch": 1.0795064241190688, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.536713600158691, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8652819395065308, "num_tokens": 323650983.0, "step": 8486 }, { "epoch": 1.0796336343976594, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.69296646118164, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8585065007209778, "num_tokens": 323690155.0, "step": 8487 }, { "epoch": 1.0797608446762499, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.536456108093262, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.879345178604126, "num_tokens": 323732212.0, "step": 8488 }, { "epoch": 1.0798880549548404, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.623022079467773, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8742655515670776, "num_tokens": 323773400.0, "step": 8489 }, { "epoch": 1.080015265233431, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.561318397521973, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8736952543258667, "num_tokens": 323812753.0, "step": 8490 }, { "epoch": 1.0801424755120215, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.566285133361816, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8687039613723755, "num_tokens": 323855835.0, "step": 8491 }, { "epoch": 1.0802696857906118, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.619119644165039, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8793304562568665, "num_tokens": 323896065.0, "step": 8492 }, { "epoch": 1.0803968960692023, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.545273780822754, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8695427179336548, "num_tokens": 323934351.0, "step": 8493 }, { "epoch": 1.0805241063477928, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.560845375061035, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8566499948501587, "num_tokens": 323968878.0, "step": 8494 }, { "epoch": 1.0806513166263834, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.613054275512695, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8652836680412292, "num_tokens": 324004838.0, "step": 8495 }, { "epoch": 1.080778526904974, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.597804069519043, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8824501037597656, "num_tokens": 324040851.0, "step": 8496 }, { "epoch": 1.0809057371835644, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.590502738952637, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8539042472839355, "num_tokens": 324084675.0, "step": 8497 }, { "epoch": 1.081032947462155, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.606306076049805, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8723114132881165, "num_tokens": 324122136.0, "step": 8498 }, { "epoch": 1.0811601577407455, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.586503982543945, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8654214143753052, "num_tokens": 324153142.0, "step": 8499 }, { "epoch": 1.081287368019336, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.592041015625, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8468518853187561, "num_tokens": 324187244.0, "step": 8500 }, { "epoch": 1.0814145782979265, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.632965087890625, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8686407804489136, "num_tokens": 324225644.0, "step": 8501 }, { "epoch": 1.081541788576517, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.528006553649902, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8747674822807312, "num_tokens": 324270394.0, "step": 8502 }, { "epoch": 1.0816689988551076, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.616008758544922, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8774823546409607, "num_tokens": 324311244.0, "step": 8503 }, { "epoch": 1.0817962091336981, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.568113327026367, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8488322496414185, "num_tokens": 324350821.0, "step": 8504 }, { "epoch": 1.0819234194122884, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.579534530639648, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8654572367668152, "num_tokens": 324388255.0, "step": 8505 }, { "epoch": 1.082050629690879, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.59911060333252, "learning_rate": 1e-06, "loss": 0.4819, "mean_token_accuracy": 0.8413835763931274, "num_tokens": 324430068.0, "step": 8506 }, { "epoch": 1.0821778399694695, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.613057136535645, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8737643361091614, "num_tokens": 324465341.0, "step": 8507 }, { "epoch": 1.08230505024806, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.578619956970215, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8743882775306702, "num_tokens": 324502199.0, "step": 8508 }, { "epoch": 1.0824322605266505, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.57776165008545, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8680754899978638, "num_tokens": 324535898.0, "step": 8509 }, { "epoch": 1.082559470805241, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.659782409667969, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8566113710403442, "num_tokens": 324574404.0, "step": 8510 }, { "epoch": 1.0826866810838316, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.551187515258789, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8655045032501221, "num_tokens": 324611743.0, "step": 8511 }, { "epoch": 1.0828138913624221, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.602262496948242, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8573048710823059, "num_tokens": 324653126.0, "step": 8512 }, { "epoch": 1.0829411016410126, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.644986152648926, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8739019632339478, "num_tokens": 324692957.0, "step": 8513 }, { "epoch": 1.0830683119196032, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.54805850982666, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8617321252822876, "num_tokens": 324737214.0, "step": 8514 }, { "epoch": 1.0831955221981937, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.58668327331543, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8772147297859192, "num_tokens": 324773098.0, "step": 8515 }, { "epoch": 1.083322732476784, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.58867073059082, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8560251593589783, "num_tokens": 324815846.0, "step": 8516 }, { "epoch": 1.0834499427553745, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.560945510864258, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8624347448348999, "num_tokens": 324855912.0, "step": 8517 }, { "epoch": 1.083577153033965, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.660289764404297, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8607094287872314, "num_tokens": 324893090.0, "step": 8518 }, { "epoch": 1.0837043633125556, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.616178512573242, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8664981722831726, "num_tokens": 324929779.0, "step": 8519 }, { "epoch": 1.0838315735911461, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.595909118652344, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8714699149131775, "num_tokens": 324961349.0, "step": 8520 }, { "epoch": 1.0839587838697367, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.64232349395752, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.866597056388855, "num_tokens": 324998688.0, "step": 8521 }, { "epoch": 1.0840859941483272, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.597394943237305, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8596581220626831, "num_tokens": 325040601.0, "step": 8522 }, { "epoch": 1.0842132044269177, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.59193229675293, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8708223700523376, "num_tokens": 325081729.0, "step": 8523 }, { "epoch": 1.0843404147055082, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.58443832397461, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8553754091262817, "num_tokens": 325118886.0, "step": 8524 }, { "epoch": 1.0844676249840988, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.572932243347168, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8841856122016907, "num_tokens": 325156417.0, "step": 8525 }, { "epoch": 1.0845948352626893, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.525932312011719, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8787416219711304, "num_tokens": 325200902.0, "step": 8526 }, { "epoch": 1.0847220455412798, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.59897232055664, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8682732582092285, "num_tokens": 325233362.0, "step": 8527 }, { "epoch": 1.0848492558198704, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.506848335266113, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8655843138694763, "num_tokens": 325274646.0, "step": 8528 }, { "epoch": 1.0849764660984607, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.621476173400879, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8820284605026245, "num_tokens": 325304735.0, "step": 8529 }, { "epoch": 1.0851036763770512, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.615930557250977, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8699626326560974, "num_tokens": 325339833.0, "step": 8530 }, { "epoch": 1.0852308866556417, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.637592315673828, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8546794056892395, "num_tokens": 325373838.0, "step": 8531 }, { "epoch": 1.0853580969342322, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.592065811157227, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.886989176273346, "num_tokens": 325412913.0, "step": 8532 }, { "epoch": 1.0854853072128228, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.547689437866211, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8719993829727173, "num_tokens": 325451830.0, "step": 8533 }, { "epoch": 1.0856125174914133, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.63850212097168, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8487350344657898, "num_tokens": 325487162.0, "step": 8534 }, { "epoch": 1.0857397277700038, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.575331687927246, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.876097559928894, "num_tokens": 325525906.0, "step": 8535 }, { "epoch": 1.0858669380485944, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.626111030578613, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8786042332649231, "num_tokens": 325559925.0, "step": 8536 }, { "epoch": 1.0859941483271849, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.614275932312012, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8664725422859192, "num_tokens": 325599980.0, "step": 8537 }, { "epoch": 1.0861213586057754, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.603193283081055, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8510335683822632, "num_tokens": 325631745.0, "step": 8538 }, { "epoch": 1.086248568884366, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.594358444213867, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8633667230606079, "num_tokens": 325673583.0, "step": 8539 }, { "epoch": 1.0863757791629565, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.591446876525879, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.859288215637207, "num_tokens": 325712995.0, "step": 8540 }, { "epoch": 1.0865029894415468, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.615598678588867, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.873092770576477, "num_tokens": 325748259.0, "step": 8541 }, { "epoch": 1.0866301997201373, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.556103706359863, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8718100786209106, "num_tokens": 325787676.0, "step": 8542 }, { "epoch": 1.0867574099987278, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.646370887756348, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8422091007232666, "num_tokens": 325827819.0, "step": 8543 }, { "epoch": 1.0868846202773184, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.560276985168457, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8753761053085327, "num_tokens": 325866328.0, "step": 8544 }, { "epoch": 1.0870118305559089, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.626079559326172, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8614959120750427, "num_tokens": 325906752.0, "step": 8545 }, { "epoch": 1.0871390408344994, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.539816856384277, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8606429696083069, "num_tokens": 325943765.0, "step": 8546 }, { "epoch": 1.08726625111309, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.569918632507324, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.873525857925415, "num_tokens": 325981471.0, "step": 8547 }, { "epoch": 1.0873934613916805, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.644113540649414, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8712736964225769, "num_tokens": 326017253.0, "step": 8548 }, { "epoch": 1.087520671670271, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.631325721740723, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8684258460998535, "num_tokens": 326053484.0, "step": 8549 }, { "epoch": 1.0876478819488615, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.654343605041504, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8718526363372803, "num_tokens": 326084779.0, "step": 8550 }, { "epoch": 1.087775092227452, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.674836158752441, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8651231527328491, "num_tokens": 326120795.0, "step": 8551 }, { "epoch": 1.0879023025060426, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.634390830993652, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8646882772445679, "num_tokens": 326155991.0, "step": 8552 }, { "epoch": 1.0880295127846331, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.620512962341309, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8633944988250732, "num_tokens": 326195738.0, "step": 8553 }, { "epoch": 1.0881567230632234, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.61104965209961, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8669708967208862, "num_tokens": 326237773.0, "step": 8554 }, { "epoch": 1.088283933341814, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.704622268676758, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8812088370323181, "num_tokens": 326278589.0, "step": 8555 }, { "epoch": 1.0884111436204045, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.574273109436035, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8596423864364624, "num_tokens": 326316862.0, "step": 8556 }, { "epoch": 1.088538353898995, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.720650672912598, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8634887933731079, "num_tokens": 326353892.0, "step": 8557 }, { "epoch": 1.0886655641775855, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.547715187072754, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8708527088165283, "num_tokens": 326389478.0, "step": 8558 }, { "epoch": 1.088792774456176, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.656394004821777, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8619027137756348, "num_tokens": 326428072.0, "step": 8559 }, { "epoch": 1.0889199847347666, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.620963096618652, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8690575957298279, "num_tokens": 326466374.0, "step": 8560 }, { "epoch": 1.0890471950133571, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.682843208312988, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8629608154296875, "num_tokens": 326501081.0, "step": 8561 }, { "epoch": 1.0891744052919476, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.627896308898926, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8753817081451416, "num_tokens": 326532539.0, "step": 8562 }, { "epoch": 1.0893016155705382, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.64698314666748, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8721821904182434, "num_tokens": 326569481.0, "step": 8563 }, { "epoch": 1.0894288258491287, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.644427299499512, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8644798994064331, "num_tokens": 326609746.0, "step": 8564 }, { "epoch": 1.089556036127719, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.583674430847168, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.856132447719574, "num_tokens": 326652379.0, "step": 8565 }, { "epoch": 1.0896832464063095, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.66511058807373, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8620381355285645, "num_tokens": 326687617.0, "step": 8566 }, { "epoch": 1.0898104566849, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.617749214172363, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.858293354511261, "num_tokens": 326724155.0, "step": 8567 }, { "epoch": 1.0899376669634906, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.576874732971191, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.865513265132904, "num_tokens": 326768234.0, "step": 8568 }, { "epoch": 1.0900648772420811, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.58530044555664, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8705018758773804, "num_tokens": 326810423.0, "step": 8569 }, { "epoch": 1.0901920875206716, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.604711532592773, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8644845485687256, "num_tokens": 326851834.0, "step": 8570 }, { "epoch": 1.0903192977992622, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.658745765686035, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8542488813400269, "num_tokens": 326889334.0, "step": 8571 }, { "epoch": 1.0904465080778527, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.628427505493164, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8553563952445984, "num_tokens": 326931884.0, "step": 8572 }, { "epoch": 1.0905737183564432, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.55428695678711, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8748173713684082, "num_tokens": 326969043.0, "step": 8573 }, { "epoch": 1.0907009286350338, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.736767768859863, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.866981565952301, "num_tokens": 326998369.0, "step": 8574 }, { "epoch": 1.0908281389136243, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.6158447265625, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8660068511962891, "num_tokens": 327036909.0, "step": 8575 }, { "epoch": 1.0909553491922148, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.647909164428711, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8597396016120911, "num_tokens": 327076053.0, "step": 8576 }, { "epoch": 1.0910825594708053, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.603233337402344, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8729342222213745, "num_tokens": 327112285.0, "step": 8577 }, { "epoch": 1.0912097697493957, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.697559356689453, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.872205913066864, "num_tokens": 327151591.0, "step": 8578 }, { "epoch": 1.0913369800279862, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.627304077148438, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8507736921310425, "num_tokens": 327193600.0, "step": 8579 }, { "epoch": 1.0914641903065767, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.647655487060547, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8570617437362671, "num_tokens": 327230249.0, "step": 8580 }, { "epoch": 1.0915914005851672, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.619755744934082, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8786853551864624, "num_tokens": 327270043.0, "step": 8581 }, { "epoch": 1.0917186108637578, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.603494644165039, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8691220879554749, "num_tokens": 327309396.0, "step": 8582 }, { "epoch": 1.0918458211423483, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.66226577758789, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8743903040885925, "num_tokens": 327349497.0, "step": 8583 }, { "epoch": 1.0919730314209388, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.642698287963867, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8586772680282593, "num_tokens": 327385648.0, "step": 8584 }, { "epoch": 1.0921002416995294, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.622657775878906, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8668777942657471, "num_tokens": 327426379.0, "step": 8585 }, { "epoch": 1.0922274519781199, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.613151550292969, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8713451027870178, "num_tokens": 327467555.0, "step": 8586 }, { "epoch": 1.0923546622567104, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.698798179626465, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8622925281524658, "num_tokens": 327505303.0, "step": 8587 }, { "epoch": 1.092481872535301, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.581080436706543, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8718982338905334, "num_tokens": 327547854.0, "step": 8588 }, { "epoch": 1.0926090828138915, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.696581840515137, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8679237365722656, "num_tokens": 327584132.0, "step": 8589 }, { "epoch": 1.0927362930924818, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.574138641357422, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8542743921279907, "num_tokens": 327629250.0, "step": 8590 }, { "epoch": 1.0928635033710723, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.655653953552246, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8582471609115601, "num_tokens": 327666988.0, "step": 8591 }, { "epoch": 1.0929907136496628, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.630532264709473, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8682675361633301, "num_tokens": 327705955.0, "step": 8592 }, { "epoch": 1.0931179239282534, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.628724098205566, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8584201335906982, "num_tokens": 327746362.0, "step": 8593 }, { "epoch": 1.0932451342068439, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.639190673828125, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.865729808807373, "num_tokens": 327789034.0, "step": 8594 }, { "epoch": 1.0933723444854344, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.60236644744873, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8556211590766907, "num_tokens": 327824203.0, "step": 8595 }, { "epoch": 1.093499554764025, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.67327880859375, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.867477536201477, "num_tokens": 327860340.0, "step": 8596 }, { "epoch": 1.0936267650426155, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.61433219909668, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8592722415924072, "num_tokens": 327901709.0, "step": 8597 }, { "epoch": 1.093753975321206, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.622257232666016, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.867359459400177, "num_tokens": 327940327.0, "step": 8598 }, { "epoch": 1.0938811855997965, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.656194686889648, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8713048696517944, "num_tokens": 327977966.0, "step": 8599 }, { "epoch": 1.094008395878387, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.58212661743164, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8619276285171509, "num_tokens": 328021143.0, "step": 8600 }, { "epoch": 1.0941356061569776, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.656747817993164, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8715877532958984, "num_tokens": 328060792.0, "step": 8601 }, { "epoch": 1.094262816435568, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.614502906799316, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8667744398117065, "num_tokens": 328099698.0, "step": 8602 }, { "epoch": 1.0943900267141584, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.635889053344727, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8698346614837646, "num_tokens": 328141075.0, "step": 8603 }, { "epoch": 1.094517236992749, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.593842506408691, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8595211505889893, "num_tokens": 328178945.0, "step": 8604 }, { "epoch": 1.0946444472713395, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.618839263916016, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8397979736328125, "num_tokens": 328214423.0, "step": 8605 }, { "epoch": 1.09477165754993, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.599264144897461, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8666318655014038, "num_tokens": 328248640.0, "step": 8606 }, { "epoch": 1.0948988678285205, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.63125991821289, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8456435203552246, "num_tokens": 328283016.0, "step": 8607 }, { "epoch": 1.095026078107111, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.581989288330078, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8695693612098694, "num_tokens": 328322377.0, "step": 8608 }, { "epoch": 1.0951532883857016, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.657110214233398, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8673710227012634, "num_tokens": 328359409.0, "step": 8609 }, { "epoch": 1.0952804986642921, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.553117752075195, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8704972267150879, "num_tokens": 328401257.0, "step": 8610 }, { "epoch": 1.0954077089428826, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.603437423706055, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.868894636631012, "num_tokens": 328435886.0, "step": 8611 }, { "epoch": 1.0955349192214732, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.58479118347168, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8561156988143921, "num_tokens": 328472091.0, "step": 8612 }, { "epoch": 1.0956621295000637, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.57817554473877, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.858116090297699, "num_tokens": 328509453.0, "step": 8613 }, { "epoch": 1.095789339778654, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.610892295837402, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8666847944259644, "num_tokens": 328545590.0, "step": 8614 }, { "epoch": 1.0959165500572445, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.615443229675293, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8558437824249268, "num_tokens": 328583093.0, "step": 8615 }, { "epoch": 1.096043760335835, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.603285789489746, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8490488529205322, "num_tokens": 328625312.0, "step": 8616 }, { "epoch": 1.0961709706144256, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.652888298034668, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8576264381408691, "num_tokens": 328655392.0, "step": 8617 }, { "epoch": 1.0962981808930161, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.595930099487305, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.850242018699646, "num_tokens": 328693686.0, "step": 8618 }, { "epoch": 1.0964253911716066, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.618026733398438, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8767696619033813, "num_tokens": 328728051.0, "step": 8619 }, { "epoch": 1.0965526014501972, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.624917030334473, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8557316660881042, "num_tokens": 328768091.0, "step": 8620 }, { "epoch": 1.0966798117287877, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.613822937011719, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8629941940307617, "num_tokens": 328803851.0, "step": 8621 }, { "epoch": 1.0968070220073782, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.61372184753418, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8767416477203369, "num_tokens": 328837429.0, "step": 8622 }, { "epoch": 1.0969342322859688, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.635786056518555, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8651789426803589, "num_tokens": 328874264.0, "step": 8623 }, { "epoch": 1.0970614425645593, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.642572402954102, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8685424327850342, "num_tokens": 328911627.0, "step": 8624 }, { "epoch": 1.0971886528431498, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.62294864654541, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.858705997467041, "num_tokens": 328944471.0, "step": 8625 }, { "epoch": 1.0973158631217403, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.613476753234863, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8734344840049744, "num_tokens": 328983826.0, "step": 8626 }, { "epoch": 1.0974430734003306, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.633625984191895, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8667889833450317, "num_tokens": 329019385.0, "step": 8627 }, { "epoch": 1.0975702836789212, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.59716510772705, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8527814149856567, "num_tokens": 329065179.0, "step": 8628 }, { "epoch": 1.0976974939575117, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.587265968322754, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8721361756324768, "num_tokens": 329101077.0, "step": 8629 }, { "epoch": 1.0978247042361022, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.661345481872559, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8710272312164307, "num_tokens": 329137538.0, "step": 8630 }, { "epoch": 1.0979519145146928, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.605234146118164, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8803725242614746, "num_tokens": 329171952.0, "step": 8631 }, { "epoch": 1.0980791247932833, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.637508392333984, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8737658858299255, "num_tokens": 329214956.0, "step": 8632 }, { "epoch": 1.0982063350718738, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.619046211242676, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8561661243438721, "num_tokens": 329254228.0, "step": 8633 }, { "epoch": 1.0983335453504643, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.661873817443848, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8672266602516174, "num_tokens": 329288337.0, "step": 8634 }, { "epoch": 1.0984607556290549, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.584868431091309, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8592520356178284, "num_tokens": 329329118.0, "step": 8635 }, { "epoch": 1.0985879659076454, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.64106559753418, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8517920970916748, "num_tokens": 329366301.0, "step": 8636 }, { "epoch": 1.098715176186236, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.662801742553711, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8553358316421509, "num_tokens": 329401612.0, "step": 8637 }, { "epoch": 1.0988423864648265, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.586834907531738, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8676574230194092, "num_tokens": 329440403.0, "step": 8638 }, { "epoch": 1.0989695967434168, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.589357376098633, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8757275938987732, "num_tokens": 329478699.0, "step": 8639 }, { "epoch": 1.0990968070220073, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.637426376342773, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8641368746757507, "num_tokens": 329517860.0, "step": 8640 }, { "epoch": 1.0992240173005978, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.601698875427246, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8715401291847229, "num_tokens": 329555385.0, "step": 8641 }, { "epoch": 1.0993512275791884, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.595884323120117, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8630290031433105, "num_tokens": 329594775.0, "step": 8642 }, { "epoch": 1.0994784378577789, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.616511344909668, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8706645965576172, "num_tokens": 329629872.0, "step": 8643 }, { "epoch": 1.0996056481363694, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.623896598815918, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8606123328208923, "num_tokens": 329665263.0, "step": 8644 }, { "epoch": 1.09973285841496, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.584741592407227, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8566380143165588, "num_tokens": 329705853.0, "step": 8645 }, { "epoch": 1.0998600686935505, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.606247901916504, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8727238178253174, "num_tokens": 329741690.0, "step": 8646 }, { "epoch": 1.099987278972141, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.667802810668945, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8579403162002563, "num_tokens": 329781548.0, "step": 8647 }, { "epoch": 1.1001144892507315, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.610190391540527, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8729429244995117, "num_tokens": 329822304.0, "step": 8648 }, { "epoch": 1.100241699529322, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.681472778320312, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8670105934143066, "num_tokens": 329856065.0, "step": 8649 }, { "epoch": 1.1003689098079126, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.685137748718262, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8504502773284912, "num_tokens": 329895002.0, "step": 8650 }, { "epoch": 1.100496120086503, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.563372611999512, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8604644536972046, "num_tokens": 329938810.0, "step": 8651 }, { "epoch": 1.1006233303650934, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.640898704528809, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8788549900054932, "num_tokens": 329976513.0, "step": 8652 }, { "epoch": 1.100750540643684, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.632643699645996, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8730994462966919, "num_tokens": 330014013.0, "step": 8653 }, { "epoch": 1.1008777509222745, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.64002799987793, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8613345623016357, "num_tokens": 330052577.0, "step": 8654 }, { "epoch": 1.101004961200865, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.661178588867188, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.863304615020752, "num_tokens": 330091385.0, "step": 8655 }, { "epoch": 1.1011321714794555, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.646498680114746, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8750005960464478, "num_tokens": 330130285.0, "step": 8656 }, { "epoch": 1.101259381758046, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.693412780761719, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8527799844741821, "num_tokens": 330166823.0, "step": 8657 }, { "epoch": 1.1013865920366366, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.671043395996094, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8667802214622498, "num_tokens": 330200655.0, "step": 8658 }, { "epoch": 1.101513802315227, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.57453441619873, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8643745183944702, "num_tokens": 330243656.0, "step": 8659 }, { "epoch": 1.1016410125938176, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.651211738586426, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8571613430976868, "num_tokens": 330287200.0, "step": 8660 }, { "epoch": 1.1017682228724082, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.611727714538574, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8683356046676636, "num_tokens": 330324794.0, "step": 8661 }, { "epoch": 1.1018954331509987, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.692200660705566, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8670870065689087, "num_tokens": 330365297.0, "step": 8662 }, { "epoch": 1.102022643429589, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.612204551696777, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8832341432571411, "num_tokens": 330408046.0, "step": 8663 }, { "epoch": 1.1021498537081795, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.64408016204834, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8613875508308411, "num_tokens": 330447815.0, "step": 8664 }, { "epoch": 1.10227706398677, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.625411987304688, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8558242321014404, "num_tokens": 330487796.0, "step": 8665 }, { "epoch": 1.1024042742653606, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.592415809631348, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8771286010742188, "num_tokens": 330531155.0, "step": 8666 }, { "epoch": 1.1025314845439511, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.67856216430664, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8750765323638916, "num_tokens": 330568611.0, "step": 8667 }, { "epoch": 1.1026586948225416, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.660128593444824, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8607538938522339, "num_tokens": 330608085.0, "step": 8668 }, { "epoch": 1.1027859051011322, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.640464782714844, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8489230871200562, "num_tokens": 330644776.0, "step": 8669 }, { "epoch": 1.1029131153797227, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.656364440917969, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8557659387588501, "num_tokens": 330686401.0, "step": 8670 }, { "epoch": 1.1030403256583132, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.583379745483398, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8705929517745972, "num_tokens": 330723115.0, "step": 8671 }, { "epoch": 1.1031675359369038, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.60796070098877, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8812461495399475, "num_tokens": 330761170.0, "step": 8672 }, { "epoch": 1.1032947462154943, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.677473068237305, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8661065101623535, "num_tokens": 330790230.0, "step": 8673 }, { "epoch": 1.1034219564940848, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.592326164245605, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8765313625335693, "num_tokens": 330826178.0, "step": 8674 }, { "epoch": 1.1035491667726753, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.64277172088623, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8723374605178833, "num_tokens": 330866341.0, "step": 8675 }, { "epoch": 1.1036763770512656, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.631939888000488, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8529351353645325, "num_tokens": 330903955.0, "step": 8676 }, { "epoch": 1.1038035873298562, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.647567749023438, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8622726202011108, "num_tokens": 330943024.0, "step": 8677 }, { "epoch": 1.1039307976084467, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.658599853515625, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.860654354095459, "num_tokens": 330981004.0, "step": 8678 }, { "epoch": 1.1040580078870372, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.659931182861328, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8648794293403625, "num_tokens": 331021532.0, "step": 8679 }, { "epoch": 1.1041852181656278, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.66922664642334, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.873423159122467, "num_tokens": 331058010.0, "step": 8680 }, { "epoch": 1.1043124284442183, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.64733600616455, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8743929266929626, "num_tokens": 331090713.0, "step": 8681 }, { "epoch": 1.1044396387228088, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.631013870239258, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8714109659194946, "num_tokens": 331127532.0, "step": 8682 }, { "epoch": 1.1045668490013993, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.603466987609863, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8453342914581299, "num_tokens": 331172088.0, "step": 8683 }, { "epoch": 1.1046940592799899, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.689632415771484, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8612974882125854, "num_tokens": 331213419.0, "step": 8684 }, { "epoch": 1.1048212695585804, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.644383430480957, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8657350540161133, "num_tokens": 331250102.0, "step": 8685 }, { "epoch": 1.104948479837171, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.622188568115234, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8672603964805603, "num_tokens": 331284597.0, "step": 8686 }, { "epoch": 1.1050756901157615, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.6835355758667, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8576626777648926, "num_tokens": 331330688.0, "step": 8687 }, { "epoch": 1.1052029003943518, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.629475593566895, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8678746819496155, "num_tokens": 331368777.0, "step": 8688 }, { "epoch": 1.1053301106729423, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.621537208557129, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8699914216995239, "num_tokens": 331414369.0, "step": 8689 }, { "epoch": 1.1054573209515328, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.669832229614258, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8715298175811768, "num_tokens": 331452788.0, "step": 8690 }, { "epoch": 1.1055845312301233, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.667037963867188, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.855148196220398, "num_tokens": 331489575.0, "step": 8691 }, { "epoch": 1.1057117415087139, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.61364459991455, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8829615116119385, "num_tokens": 331529129.0, "step": 8692 }, { "epoch": 1.1058389517873044, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.711808204650879, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8613398671150208, "num_tokens": 331564989.0, "step": 8693 }, { "epoch": 1.105966162065895, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.662484169006348, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8668151497840881, "num_tokens": 331602561.0, "step": 8694 }, { "epoch": 1.1060933723444855, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.700167655944824, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8671208620071411, "num_tokens": 331646902.0, "step": 8695 }, { "epoch": 1.106220582623076, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.73851490020752, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8665748834609985, "num_tokens": 331682347.0, "step": 8696 }, { "epoch": 1.1063477929016665, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.609938621520996, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8579665422439575, "num_tokens": 331725813.0, "step": 8697 }, { "epoch": 1.106475003180257, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.731813430786133, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8454118967056274, "num_tokens": 331767856.0, "step": 8698 }, { "epoch": 1.1066022134588476, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.693778991699219, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8746092319488525, "num_tokens": 331804238.0, "step": 8699 }, { "epoch": 1.106729423737438, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.750643730163574, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8676939606666565, "num_tokens": 331839963.0, "step": 8700 }, { "epoch": 1.1068566340160284, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.645980834960938, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.86894690990448, "num_tokens": 331881563.0, "step": 8701 }, { "epoch": 1.106983844294619, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.661030769348145, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8629411458969116, "num_tokens": 331925540.0, "step": 8702 }, { "epoch": 1.1071110545732095, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.715113639831543, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8734129071235657, "num_tokens": 331963171.0, "step": 8703 }, { "epoch": 1.1072382648518, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.713165283203125, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8861690163612366, "num_tokens": 331998310.0, "step": 8704 }, { "epoch": 1.1073654751303905, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.645989418029785, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8715966939926147, "num_tokens": 332034181.0, "step": 8705 }, { "epoch": 1.107492685408981, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.640886306762695, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8748423457145691, "num_tokens": 332075368.0, "step": 8706 }, { "epoch": 1.1076198956875716, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.641499519348145, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8767341375350952, "num_tokens": 332113281.0, "step": 8707 }, { "epoch": 1.107747105966162, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.722648620605469, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8581317067146301, "num_tokens": 332154206.0, "step": 8708 }, { "epoch": 1.1078743162447526, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.752236366271973, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8511652946472168, "num_tokens": 332195409.0, "step": 8709 }, { "epoch": 1.1080015265233432, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.605493545532227, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8748266696929932, "num_tokens": 332233836.0, "step": 8710 }, { "epoch": 1.1081287368019337, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.658400535583496, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.871060848236084, "num_tokens": 332274167.0, "step": 8711 }, { "epoch": 1.108255947080524, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.75089168548584, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8631625175476074, "num_tokens": 332315856.0, "step": 8712 }, { "epoch": 1.1083831573591145, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.678976058959961, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8657412528991699, "num_tokens": 332366365.0, "step": 8713 }, { "epoch": 1.108510367637705, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.694374084472656, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8633446097373962, "num_tokens": 332400411.0, "step": 8714 }, { "epoch": 1.1086375779162956, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.71631908416748, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8685165643692017, "num_tokens": 332441959.0, "step": 8715 }, { "epoch": 1.108764788194886, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.665865898132324, "learning_rate": 1e-06, "loss": 0.4749, "mean_token_accuracy": 0.8528587818145752, "num_tokens": 332479545.0, "step": 8716 }, { "epoch": 1.1088919984734766, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.713083267211914, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8771199584007263, "num_tokens": 332516240.0, "step": 8717 }, { "epoch": 1.1090192087520672, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.71166706085205, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8711621165275574, "num_tokens": 332553488.0, "step": 8718 }, { "epoch": 1.1091464190306577, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.706236839294434, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.856904149055481, "num_tokens": 332588646.0, "step": 8719 }, { "epoch": 1.1092736293092482, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.70839786529541, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8559600114822388, "num_tokens": 332629103.0, "step": 8720 }, { "epoch": 1.1094008395878387, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.685697555541992, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8720016479492188, "num_tokens": 332664820.0, "step": 8721 }, { "epoch": 1.1095280498664293, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.68504810333252, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8679999113082886, "num_tokens": 332699772.0, "step": 8722 }, { "epoch": 1.1096552601450198, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.615553855895996, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8741421699523926, "num_tokens": 332733478.0, "step": 8723 }, { "epoch": 1.1097824704236103, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.757108688354492, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.86133873462677, "num_tokens": 332774733.0, "step": 8724 }, { "epoch": 1.1099096807022006, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.64606761932373, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.842717170715332, "num_tokens": 332812970.0, "step": 8725 }, { "epoch": 1.1100368909807912, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.667950630187988, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8619880676269531, "num_tokens": 332854974.0, "step": 8726 }, { "epoch": 1.1101641012593817, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.670026779174805, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8525251746177673, "num_tokens": 332886462.0, "step": 8727 }, { "epoch": 1.1102913115379722, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.578557014465332, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8622013926506042, "num_tokens": 332928614.0, "step": 8728 }, { "epoch": 1.1104185218165628, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.696931838989258, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8673375248908997, "num_tokens": 332962447.0, "step": 8729 }, { "epoch": 1.1105457320951533, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.633842468261719, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8574023246765137, "num_tokens": 332991395.0, "step": 8730 }, { "epoch": 1.1106729423737438, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.65950870513916, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8584555387496948, "num_tokens": 333027852.0, "step": 8731 }, { "epoch": 1.1108001526523343, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.691747665405273, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8683390617370605, "num_tokens": 333063337.0, "step": 8732 }, { "epoch": 1.1109273629309249, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.670490264892578, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8573013544082642, "num_tokens": 333094632.0, "step": 8733 }, { "epoch": 1.1110545732095154, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.685760498046875, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8795884251594543, "num_tokens": 333134071.0, "step": 8734 }, { "epoch": 1.111181783488106, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.683612823486328, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8720426559448242, "num_tokens": 333171288.0, "step": 8735 }, { "epoch": 1.1113089937666965, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.651647567749023, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8734517097473145, "num_tokens": 333212111.0, "step": 8736 }, { "epoch": 1.1114362040452868, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.688064575195312, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.871819019317627, "num_tokens": 333247557.0, "step": 8737 }, { "epoch": 1.1115634143238773, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.724964141845703, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8652154803276062, "num_tokens": 333278705.0, "step": 8738 }, { "epoch": 1.1116906246024678, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.669442176818848, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8681203126907349, "num_tokens": 333316493.0, "step": 8739 }, { "epoch": 1.1118178348810583, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.671940803527832, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8688777089118958, "num_tokens": 333360384.0, "step": 8740 }, { "epoch": 1.1119450451596489, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.691938400268555, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8662865161895752, "num_tokens": 333395568.0, "step": 8741 }, { "epoch": 1.1120722554382394, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.730476379394531, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8704081773757935, "num_tokens": 333429267.0, "step": 8742 }, { "epoch": 1.11219946571683, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.661674499511719, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8722571730613708, "num_tokens": 333469908.0, "step": 8743 }, { "epoch": 1.1123266759954205, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.629218101501465, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.869907021522522, "num_tokens": 333507865.0, "step": 8744 }, { "epoch": 1.112453886274011, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.69294548034668, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8768346309661865, "num_tokens": 333544011.0, "step": 8745 }, { "epoch": 1.1125810965526015, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.669876098632812, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8680092096328735, "num_tokens": 333581949.0, "step": 8746 }, { "epoch": 1.112708306831192, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.632500648498535, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8744712471961975, "num_tokens": 333615994.0, "step": 8747 }, { "epoch": 1.1128355171097826, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.711029052734375, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8823732733726501, "num_tokens": 333654454.0, "step": 8748 }, { "epoch": 1.112962727388373, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.629395484924316, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8585854768753052, "num_tokens": 333700206.0, "step": 8749 }, { "epoch": 1.1130899376669634, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.676177978515625, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8699370622634888, "num_tokens": 333737538.0, "step": 8750 }, { "epoch": 1.113217147945554, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.670741081237793, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.868637204170227, "num_tokens": 333771237.0, "step": 8751 }, { "epoch": 1.1133443582241445, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.69780445098877, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8628835678100586, "num_tokens": 333806522.0, "step": 8752 }, { "epoch": 1.113471568502735, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.61824893951416, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8660162091255188, "num_tokens": 333846656.0, "step": 8753 }, { "epoch": 1.1135987787813255, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.634031295776367, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8577777147293091, "num_tokens": 333886294.0, "step": 8754 }, { "epoch": 1.113725989059916, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.666288375854492, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8639021515846252, "num_tokens": 333923868.0, "step": 8755 }, { "epoch": 1.1138531993385066, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.732155799865723, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8703096508979797, "num_tokens": 333957187.0, "step": 8756 }, { "epoch": 1.113980409617097, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.580707550048828, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8800398707389832, "num_tokens": 333999061.0, "step": 8757 }, { "epoch": 1.1141076198956876, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.701835632324219, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8608651161193848, "num_tokens": 334043595.0, "step": 8758 }, { "epoch": 1.1142348301742782, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.6231050491333, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8708359003067017, "num_tokens": 334078487.0, "step": 8759 }, { "epoch": 1.1143620404528687, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.611700057983398, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8747488260269165, "num_tokens": 334119610.0, "step": 8760 }, { "epoch": 1.114489250731459, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.660645484924316, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8623300194740295, "num_tokens": 334160149.0, "step": 8761 }, { "epoch": 1.1146164610100495, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.667282104492188, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8674583435058594, "num_tokens": 334198667.0, "step": 8762 }, { "epoch": 1.11474367128864, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.66048526763916, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8682839274406433, "num_tokens": 334233838.0, "step": 8763 }, { "epoch": 1.1148708815672306, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.65360164642334, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8590089678764343, "num_tokens": 334275703.0, "step": 8764 }, { "epoch": 1.114998091845821, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.671486854553223, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8780014514923096, "num_tokens": 334311119.0, "step": 8765 }, { "epoch": 1.1151253021244116, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.743354797363281, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8598389625549316, "num_tokens": 334348298.0, "step": 8766 }, { "epoch": 1.1152525124030022, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.669499397277832, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8540971279144287, "num_tokens": 334386986.0, "step": 8767 }, { "epoch": 1.1153797226815927, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.682101249694824, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8704228401184082, "num_tokens": 334423263.0, "step": 8768 }, { "epoch": 1.1155069329601832, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.6251859664917, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8831859230995178, "num_tokens": 334460363.0, "step": 8769 }, { "epoch": 1.1156341432387737, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.71340560913086, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8704656362533569, "num_tokens": 334501883.0, "step": 8770 }, { "epoch": 1.1157613535173643, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.672449111938477, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8661428689956665, "num_tokens": 334536242.0, "step": 8771 }, { "epoch": 1.1158885637959548, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.676950454711914, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8780006766319275, "num_tokens": 334569685.0, "step": 8772 }, { "epoch": 1.1160157740745453, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.667779922485352, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8688851594924927, "num_tokens": 334604946.0, "step": 8773 }, { "epoch": 1.1161429843531356, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.743035316467285, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8728209137916565, "num_tokens": 334645495.0, "step": 8774 }, { "epoch": 1.1162701946317262, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.70767593383789, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8612854480743408, "num_tokens": 334684239.0, "step": 8775 }, { "epoch": 1.1163974049103167, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.67173957824707, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8729079961776733, "num_tokens": 334713799.0, "step": 8776 }, { "epoch": 1.1165246151889072, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.644498825073242, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8868699073791504, "num_tokens": 334753374.0, "step": 8777 }, { "epoch": 1.1166518254674977, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.661084175109863, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8679596781730652, "num_tokens": 334792867.0, "step": 8778 }, { "epoch": 1.1167790357460883, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.740723609924316, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8636773824691772, "num_tokens": 334832664.0, "step": 8779 }, { "epoch": 1.1169062460246788, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.693531036376953, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8600959181785583, "num_tokens": 334876763.0, "step": 8780 }, { "epoch": 1.1170334563032693, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.726140022277832, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8677362203598022, "num_tokens": 334909431.0, "step": 8781 }, { "epoch": 1.1171606665818599, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.702550888061523, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8649833798408508, "num_tokens": 334941939.0, "step": 8782 }, { "epoch": 1.1172878768604504, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.692760467529297, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8759557008743286, "num_tokens": 334981338.0, "step": 8783 }, { "epoch": 1.117415087139041, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.642143249511719, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8674641847610474, "num_tokens": 335020660.0, "step": 8784 }, { "epoch": 1.1175422974176314, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.781375885009766, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8696619272232056, "num_tokens": 335057377.0, "step": 8785 }, { "epoch": 1.1176695076962218, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.66722297668457, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8572713732719421, "num_tokens": 335092912.0, "step": 8786 }, { "epoch": 1.1177967179748123, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.731619834899902, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8820437788963318, "num_tokens": 335129157.0, "step": 8787 }, { "epoch": 1.1179239282534028, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.681922912597656, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8678800463676453, "num_tokens": 335168065.0, "step": 8788 }, { "epoch": 1.1180511385319933, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.715450286865234, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8652455806732178, "num_tokens": 335218876.0, "step": 8789 }, { "epoch": 1.1181783488105839, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.629727363586426, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8765007257461548, "num_tokens": 335256300.0, "step": 8790 }, { "epoch": 1.1183055590891744, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.705623626708984, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8764960765838623, "num_tokens": 335291035.0, "step": 8791 }, { "epoch": 1.118432769367765, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.66117000579834, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8780250549316406, "num_tokens": 335330092.0, "step": 8792 }, { "epoch": 1.1185599796463555, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.705948829650879, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8538830280303955, "num_tokens": 335367146.0, "step": 8793 }, { "epoch": 1.118687189924946, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.67895793914795, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.86729896068573, "num_tokens": 335405069.0, "step": 8794 }, { "epoch": 1.1188144002035365, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.729121208190918, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8583487272262573, "num_tokens": 335443779.0, "step": 8795 }, { "epoch": 1.118941610482127, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.719663619995117, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8597515821456909, "num_tokens": 335485978.0, "step": 8796 }, { "epoch": 1.1190688207607176, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.651183128356934, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8713935613632202, "num_tokens": 335530375.0, "step": 8797 }, { "epoch": 1.119196031039308, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.710577964782715, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8719385266304016, "num_tokens": 335569109.0, "step": 8798 }, { "epoch": 1.1193232413178984, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.761133193969727, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8540340662002563, "num_tokens": 335599283.0, "step": 8799 }, { "epoch": 1.119450451596489, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.683639526367188, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8744717836380005, "num_tokens": 335634792.0, "step": 8800 }, { "epoch": 1.1195776618750795, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.72532844543457, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8802322149276733, "num_tokens": 335672869.0, "step": 8801 }, { "epoch": 1.11970487215367, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.70292854309082, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8724839687347412, "num_tokens": 335714996.0, "step": 8802 }, { "epoch": 1.1198320824322605, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.698172569274902, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8586644530296326, "num_tokens": 335756884.0, "step": 8803 }, { "epoch": 1.119959292710851, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.614645957946777, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8765577077865601, "num_tokens": 335793935.0, "step": 8804 }, { "epoch": 1.1200865029894416, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.691946029663086, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8790273666381836, "num_tokens": 335834169.0, "step": 8805 }, { "epoch": 1.120213713268032, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.64547061920166, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8859167695045471, "num_tokens": 335872302.0, "step": 8806 }, { "epoch": 1.1203409235466226, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.687056541442871, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8670222759246826, "num_tokens": 335908074.0, "step": 8807 }, { "epoch": 1.1204681338252132, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.721960067749023, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8787079453468323, "num_tokens": 335946268.0, "step": 8808 }, { "epoch": 1.1205953441038037, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.69047737121582, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8680490851402283, "num_tokens": 335985364.0, "step": 8809 }, { "epoch": 1.120722554382394, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.722818374633789, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8707802295684814, "num_tokens": 336025300.0, "step": 8810 }, { "epoch": 1.1208497646609845, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.690472602844238, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8718116283416748, "num_tokens": 336062105.0, "step": 8811 }, { "epoch": 1.120976974939575, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.656982421875, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8566556572914124, "num_tokens": 336102944.0, "step": 8812 }, { "epoch": 1.1211041852181656, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.69441032409668, "learning_rate": 1e-06, "loss": 0.3292, "mean_token_accuracy": 0.8930978775024414, "num_tokens": 336142964.0, "step": 8813 }, { "epoch": 1.121231395496756, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.682366371154785, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8706647157669067, "num_tokens": 336186962.0, "step": 8814 }, { "epoch": 1.1213586057753466, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.682027816772461, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8770638704299927, "num_tokens": 336220775.0, "step": 8815 }, { "epoch": 1.1214858160539372, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.692625045776367, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8643706440925598, "num_tokens": 336263067.0, "step": 8816 }, { "epoch": 1.1216130263325277, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.717977523803711, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8714518547058105, "num_tokens": 336308463.0, "step": 8817 }, { "epoch": 1.1217402366111182, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.609123229980469, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8806639313697815, "num_tokens": 336346965.0, "step": 8818 }, { "epoch": 1.1218674468897087, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.75271987915039, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8688541054725647, "num_tokens": 336382114.0, "step": 8819 }, { "epoch": 1.1219946571682993, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.698972702026367, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8719178438186646, "num_tokens": 336414020.0, "step": 8820 }, { "epoch": 1.1221218674468898, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.701723098754883, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8653087019920349, "num_tokens": 336452522.0, "step": 8821 }, { "epoch": 1.1222490777254803, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.698299407958984, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8780425786972046, "num_tokens": 336484593.0, "step": 8822 }, { "epoch": 1.1223762880040706, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.735220909118652, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8515226244926453, "num_tokens": 336518964.0, "step": 8823 }, { "epoch": 1.1225034982826612, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.672603607177734, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8553240299224854, "num_tokens": 336556320.0, "step": 8824 }, { "epoch": 1.1226307085612517, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.66840648651123, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8580542802810669, "num_tokens": 336595473.0, "step": 8825 }, { "epoch": 1.1227579188398422, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.763911247253418, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8674038052558899, "num_tokens": 336631793.0, "step": 8826 }, { "epoch": 1.1228851291184327, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.673991203308105, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8538230657577515, "num_tokens": 336671230.0, "step": 8827 }, { "epoch": 1.1230123393970233, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.724485397338867, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8686058521270752, "num_tokens": 336712015.0, "step": 8828 }, { "epoch": 1.1231395496756138, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.729134559631348, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8522178530693054, "num_tokens": 336741206.0, "step": 8829 }, { "epoch": 1.1232667599542043, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.64146614074707, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8598345518112183, "num_tokens": 336778838.0, "step": 8830 }, { "epoch": 1.1233939702327949, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.717804908752441, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8733911514282227, "num_tokens": 336815471.0, "step": 8831 }, { "epoch": 1.1235211805113854, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.671096801757812, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8700326681137085, "num_tokens": 336850870.0, "step": 8832 }, { "epoch": 1.123648390789976, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.699455261230469, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8725436925888062, "num_tokens": 336891779.0, "step": 8833 }, { "epoch": 1.1237756010685664, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.633825302124023, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8779213428497314, "num_tokens": 336931171.0, "step": 8834 }, { "epoch": 1.1239028113471567, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.7300443649292, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.877422571182251, "num_tokens": 336963080.0, "step": 8835 }, { "epoch": 1.1240300216257473, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.72828197479248, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8541483283042908, "num_tokens": 337000817.0, "step": 8836 }, { "epoch": 1.1241572319043378, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.673355102539062, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8616547584533691, "num_tokens": 337037249.0, "step": 8837 }, { "epoch": 1.1242844421829283, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.72103214263916, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8756085634231567, "num_tokens": 337077752.0, "step": 8838 }, { "epoch": 1.1244116524615189, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.70523738861084, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8648040890693665, "num_tokens": 337119649.0, "step": 8839 }, { "epoch": 1.1245388627401094, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.758611679077148, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8518017530441284, "num_tokens": 337154376.0, "step": 8840 }, { "epoch": 1.1246660730187, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.66445255279541, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8768178224563599, "num_tokens": 337193377.0, "step": 8841 }, { "epoch": 1.1247932832972904, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.689498901367188, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8581957221031189, "num_tokens": 337233176.0, "step": 8842 }, { "epoch": 1.124920493575881, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.736201286315918, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.852423906326294, "num_tokens": 337262995.0, "step": 8843 }, { "epoch": 1.1250477038544715, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.759166717529297, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8663489818572998, "num_tokens": 337302466.0, "step": 8844 }, { "epoch": 1.125174914133062, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.587037086486816, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8781716823577881, "num_tokens": 337341197.0, "step": 8845 }, { "epoch": 1.1253021244116526, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.714056015014648, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8688396215438843, "num_tokens": 337375602.0, "step": 8846 }, { "epoch": 1.125429334690243, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.74692440032959, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8590127825737, "num_tokens": 337406685.0, "step": 8847 }, { "epoch": 1.1255565449688334, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.672159194946289, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8631243705749512, "num_tokens": 337447364.0, "step": 8848 }, { "epoch": 1.125683755247424, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.682376861572266, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8670982718467712, "num_tokens": 337487840.0, "step": 8849 }, { "epoch": 1.1258109655260145, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.748847007751465, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8714566826820374, "num_tokens": 337525117.0, "step": 8850 }, { "epoch": 1.125938175804605, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.66444206237793, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8609802722930908, "num_tokens": 337567024.0, "step": 8851 }, { "epoch": 1.1260653860831955, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.733798027038574, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8657025098800659, "num_tokens": 337600117.0, "step": 8852 }, { "epoch": 1.126192596361786, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.730825424194336, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8526343107223511, "num_tokens": 337637188.0, "step": 8853 }, { "epoch": 1.1263198066403766, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.730469703674316, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8626229763031006, "num_tokens": 337676492.0, "step": 8854 }, { "epoch": 1.126447016918967, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.677933692932129, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8802754878997803, "num_tokens": 337713468.0, "step": 8855 }, { "epoch": 1.1265742271975576, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.78449821472168, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.853858470916748, "num_tokens": 337750944.0, "step": 8856 }, { "epoch": 1.1267014374761481, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.685599327087402, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8814948797225952, "num_tokens": 337793802.0, "step": 8857 }, { "epoch": 1.1268286477547387, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.696566581726074, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8803198337554932, "num_tokens": 337828550.0, "step": 8858 }, { "epoch": 1.126955858033329, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.750656127929688, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.876552402973175, "num_tokens": 337862168.0, "step": 8859 }, { "epoch": 1.1270830683119195, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.707947731018066, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8630794286727905, "num_tokens": 337898837.0, "step": 8860 }, { "epoch": 1.12721027859051, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.694913864135742, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.854790210723877, "num_tokens": 337939608.0, "step": 8861 }, { "epoch": 1.1273374888691006, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.681997299194336, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8662171363830566, "num_tokens": 337983412.0, "step": 8862 }, { "epoch": 1.127464699147691, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.742107391357422, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8765665292739868, "num_tokens": 338022171.0, "step": 8863 }, { "epoch": 1.1275919094262816, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.732419967651367, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8648273944854736, "num_tokens": 338058838.0, "step": 8864 }, { "epoch": 1.1277191197048722, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.737557411193848, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8637200593948364, "num_tokens": 338095080.0, "step": 8865 }, { "epoch": 1.1278463299834627, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.68692398071289, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8492130637168884, "num_tokens": 338134206.0, "step": 8866 }, { "epoch": 1.1279735402620532, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.771391868591309, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8734147548675537, "num_tokens": 338173794.0, "step": 8867 }, { "epoch": 1.1281007505406437, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.730326652526855, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8647837042808533, "num_tokens": 338206591.0, "step": 8868 }, { "epoch": 1.1282279608192343, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.809910774230957, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.879737138748169, "num_tokens": 338238728.0, "step": 8869 }, { "epoch": 1.1283551710978248, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.690868377685547, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8766346573829651, "num_tokens": 338282372.0, "step": 8870 }, { "epoch": 1.1284823813764153, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.749016761779785, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8628652095794678, "num_tokens": 338319175.0, "step": 8871 }, { "epoch": 1.1286095916550056, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.770779609680176, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8642630577087402, "num_tokens": 338357967.0, "step": 8872 }, { "epoch": 1.1287368019335962, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.754011154174805, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8739046454429626, "num_tokens": 338391431.0, "step": 8873 }, { "epoch": 1.1288640122121867, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.717382431030273, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8753658533096313, "num_tokens": 338426298.0, "step": 8874 }, { "epoch": 1.1289912224907772, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.814895629882812, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8493927121162415, "num_tokens": 338468147.0, "step": 8875 }, { "epoch": 1.1291184327693677, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.732870101928711, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8510594367980957, "num_tokens": 338517095.0, "step": 8876 }, { "epoch": 1.1292456430479583, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.745903015136719, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8745241165161133, "num_tokens": 338549925.0, "step": 8877 }, { "epoch": 1.1293728533265488, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.7883882522583, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8705486059188843, "num_tokens": 338583966.0, "step": 8878 }, { "epoch": 1.1295000636051393, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.743734359741211, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.871850311756134, "num_tokens": 338622715.0, "step": 8879 }, { "epoch": 1.1296272738837299, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.812675476074219, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8542042970657349, "num_tokens": 338659920.0, "step": 8880 }, { "epoch": 1.1297544841623204, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.782332420349121, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8627514839172363, "num_tokens": 338698183.0, "step": 8881 }, { "epoch": 1.129881694440911, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.690442085266113, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8557973504066467, "num_tokens": 338735853.0, "step": 8882 }, { "epoch": 1.1300089047195012, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.771130561828613, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8569102883338928, "num_tokens": 338771714.0, "step": 8883 }, { "epoch": 1.1301361149980917, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.700826644897461, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8830182552337646, "num_tokens": 338813823.0, "step": 8884 }, { "epoch": 1.1302633252766823, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.679859161376953, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8715528249740601, "num_tokens": 338858137.0, "step": 8885 }, { "epoch": 1.1303905355552728, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.79383659362793, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8576936721801758, "num_tokens": 338893653.0, "step": 8886 }, { "epoch": 1.1305177458338633, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.784972190856934, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8479391932487488, "num_tokens": 338931230.0, "step": 8887 }, { "epoch": 1.1306449561124539, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.723793029785156, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8669161200523376, "num_tokens": 338966921.0, "step": 8888 }, { "epoch": 1.1307721663910444, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.796472549438477, "learning_rate": 1e-06, "loss": 0.4873, "mean_token_accuracy": 0.8474625945091248, "num_tokens": 339000957.0, "step": 8889 }, { "epoch": 1.130899376669635, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.704248428344727, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8590271472930908, "num_tokens": 339035876.0, "step": 8890 }, { "epoch": 1.1310265869482254, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.756599426269531, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8683452010154724, "num_tokens": 339073700.0, "step": 8891 }, { "epoch": 1.131153797226816, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.788948059082031, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8712248802185059, "num_tokens": 339108222.0, "step": 8892 }, { "epoch": 1.1312810075054065, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.705658912658691, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8776202201843262, "num_tokens": 339145749.0, "step": 8893 }, { "epoch": 1.131408217783997, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.806415557861328, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.853750467300415, "num_tokens": 339183852.0, "step": 8894 }, { "epoch": 1.1315354280625876, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.71766471862793, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8824880123138428, "num_tokens": 339218766.0, "step": 8895 }, { "epoch": 1.131662638341178, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.705159187316895, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8676359057426453, "num_tokens": 339259932.0, "step": 8896 }, { "epoch": 1.1317898486197684, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.756270408630371, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8701587915420532, "num_tokens": 339302201.0, "step": 8897 }, { "epoch": 1.131917058898359, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.67615032196045, "learning_rate": 1e-06, "loss": 0.5218, "mean_token_accuracy": 0.8381422758102417, "num_tokens": 339342091.0, "step": 8898 }, { "epoch": 1.1320442691769494, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.798966407775879, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8622978925704956, "num_tokens": 339379951.0, "step": 8899 }, { "epoch": 1.13217147945554, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.722189903259277, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8615326881408691, "num_tokens": 339422243.0, "step": 8900 }, { "epoch": 1.1322986897341305, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.681859016418457, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8813672065734863, "num_tokens": 339464969.0, "step": 8901 }, { "epoch": 1.132425900012721, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.714686393737793, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8706748485565186, "num_tokens": 339502382.0, "step": 8902 }, { "epoch": 1.1325531102913116, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.709320068359375, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8515225648880005, "num_tokens": 339542875.0, "step": 8903 }, { "epoch": 1.132680320569902, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.734972953796387, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8543276786804199, "num_tokens": 339580776.0, "step": 8904 }, { "epoch": 1.1328075308484926, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.715197563171387, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8799604773521423, "num_tokens": 339620195.0, "step": 8905 }, { "epoch": 1.1329347411270831, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.768702507019043, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8749151229858398, "num_tokens": 339659234.0, "step": 8906 }, { "epoch": 1.1330619514056737, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.705997467041016, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8573702573776245, "num_tokens": 339703259.0, "step": 8907 }, { "epoch": 1.133189161684264, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.688814163208008, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.873793363571167, "num_tokens": 339745901.0, "step": 8908 }, { "epoch": 1.1333163719628545, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.707175254821777, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8576509952545166, "num_tokens": 339790417.0, "step": 8909 }, { "epoch": 1.133443582241445, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.729856491088867, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8597030639648438, "num_tokens": 339827224.0, "step": 8910 }, { "epoch": 1.1335707925200356, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.7395658493042, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8610284328460693, "num_tokens": 339859932.0, "step": 8911 }, { "epoch": 1.133698002798626, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.732505798339844, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8554461002349854, "num_tokens": 339902416.0, "step": 8912 }, { "epoch": 1.1338252130772166, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.709206581115723, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8757691979408264, "num_tokens": 339940495.0, "step": 8913 }, { "epoch": 1.1339524233558071, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.690961837768555, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8496528267860413, "num_tokens": 339983010.0, "step": 8914 }, { "epoch": 1.1340796336343977, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.76524543762207, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8667160272598267, "num_tokens": 340020683.0, "step": 8915 }, { "epoch": 1.1342068439129882, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.732558250427246, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8577525019645691, "num_tokens": 340063179.0, "step": 8916 }, { "epoch": 1.1343340541915787, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.742037773132324, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8762192130088806, "num_tokens": 340100166.0, "step": 8917 }, { "epoch": 1.1344612644701693, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.757935523986816, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8634563088417053, "num_tokens": 340144325.0, "step": 8918 }, { "epoch": 1.1345884747487598, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.74191665649414, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8659393787384033, "num_tokens": 340180563.0, "step": 8919 }, { "epoch": 1.1347156850273503, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.733306884765625, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8530073165893555, "num_tokens": 340224621.0, "step": 8920 }, { "epoch": 1.1348428953059406, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.779844284057617, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8718928694725037, "num_tokens": 340264810.0, "step": 8921 }, { "epoch": 1.1349701055845312, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.744041442871094, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8790918588638306, "num_tokens": 340301158.0, "step": 8922 }, { "epoch": 1.1350973158631217, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75150203704834, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8668216466903687, "num_tokens": 340339806.0, "step": 8923 }, { "epoch": 1.1352245261417122, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.777018547058105, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8716071844100952, "num_tokens": 340386412.0, "step": 8924 }, { "epoch": 1.1353517364203027, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.75318431854248, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.876013994216919, "num_tokens": 340431391.0, "step": 8925 }, { "epoch": 1.1354789466988933, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.808170318603516, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8701281547546387, "num_tokens": 340465904.0, "step": 8926 }, { "epoch": 1.1356061569774838, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.787958145141602, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8703814148902893, "num_tokens": 340504290.0, "step": 8927 }, { "epoch": 1.1357333672560743, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.806134223937988, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.861526370048523, "num_tokens": 340539100.0, "step": 8928 }, { "epoch": 1.1358605775346649, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.75043773651123, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8810917139053345, "num_tokens": 340569005.0, "step": 8929 }, { "epoch": 1.1359877878132554, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.725419998168945, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8669982552528381, "num_tokens": 340613476.0, "step": 8930 }, { "epoch": 1.136114998091846, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.783171653747559, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8588639497756958, "num_tokens": 340647982.0, "step": 8931 }, { "epoch": 1.1362422083704362, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.77157974243164, "learning_rate": 1e-06, "loss": 0.4897, "mean_token_accuracy": 0.8451712131500244, "num_tokens": 340683270.0, "step": 8932 }, { "epoch": 1.1363694186490267, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.806037902832031, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8701040744781494, "num_tokens": 340719955.0, "step": 8933 }, { "epoch": 1.1364966289276173, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.765143394470215, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8673961162567139, "num_tokens": 340753934.0, "step": 8934 }, { "epoch": 1.1366238392062078, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.735416412353516, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8728517293930054, "num_tokens": 340787545.0, "step": 8935 }, { "epoch": 1.1367510494847983, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.657859802246094, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8764021992683411, "num_tokens": 340826362.0, "step": 8936 }, { "epoch": 1.1368782597633889, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.749828338623047, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8703957200050354, "num_tokens": 340857871.0, "step": 8937 }, { "epoch": 1.1370054700419794, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.72681713104248, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8721714019775391, "num_tokens": 340897736.0, "step": 8938 }, { "epoch": 1.13713268032057, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.727338790893555, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.862983763217926, "num_tokens": 340935774.0, "step": 8939 }, { "epoch": 1.1372598905991604, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.674050331115723, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8707578778266907, "num_tokens": 340971497.0, "step": 8940 }, { "epoch": 1.137387100877751, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.696170806884766, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8680493235588074, "num_tokens": 341016209.0, "step": 8941 }, { "epoch": 1.1375143111563415, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.717533111572266, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8625195026397705, "num_tokens": 341052137.0, "step": 8942 }, { "epoch": 1.137641521434932, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.760780334472656, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8695554733276367, "num_tokens": 341083417.0, "step": 8943 }, { "epoch": 1.1377687317135226, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.660264015197754, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8702266216278076, "num_tokens": 341127798.0, "step": 8944 }, { "epoch": 1.137895941992113, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.72824478149414, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8530117273330688, "num_tokens": 341173485.0, "step": 8945 }, { "epoch": 1.1380231522707034, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.724268913269043, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8558247089385986, "num_tokens": 341212834.0, "step": 8946 }, { "epoch": 1.138150362549294, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.724308967590332, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8599323630332947, "num_tokens": 341246497.0, "step": 8947 }, { "epoch": 1.1382775728278844, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.681144714355469, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8517207503318787, "num_tokens": 341291498.0, "step": 8948 }, { "epoch": 1.138404783106475, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.733915328979492, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8603700995445251, "num_tokens": 341328104.0, "step": 8949 }, { "epoch": 1.1385319933850655, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.722380638122559, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8777061700820923, "num_tokens": 341359042.0, "step": 8950 }, { "epoch": 1.138659203663656, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.720942497253418, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8666814565658569, "num_tokens": 341403529.0, "step": 8951 }, { "epoch": 1.1387864139422466, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.702917098999023, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8736085891723633, "num_tokens": 341442509.0, "step": 8952 }, { "epoch": 1.138913624220837, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.746075630187988, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8645368814468384, "num_tokens": 341483015.0, "step": 8953 }, { "epoch": 1.1390408344994276, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.673932075500488, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.867059588432312, "num_tokens": 341519676.0, "step": 8954 }, { "epoch": 1.1391680447780181, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.80274486541748, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8667391538619995, "num_tokens": 341563483.0, "step": 8955 }, { "epoch": 1.1392952550566087, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.759320259094238, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.859540581703186, "num_tokens": 341598387.0, "step": 8956 }, { "epoch": 1.139422465335199, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.756025314331055, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8697139620780945, "num_tokens": 341640253.0, "step": 8957 }, { "epoch": 1.1395496756137895, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.749517440795898, "learning_rate": 1e-06, "loss": 0.3351, "mean_token_accuracy": 0.8939300179481506, "num_tokens": 341675788.0, "step": 8958 }, { "epoch": 1.13967688589238, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.624150276184082, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8663893938064575, "num_tokens": 341716917.0, "step": 8959 }, { "epoch": 1.1398040961709706, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.768256187438965, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8654923439025879, "num_tokens": 341758936.0, "step": 8960 }, { "epoch": 1.139931306449561, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.73412036895752, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8436393737792969, "num_tokens": 341797612.0, "step": 8961 }, { "epoch": 1.1400585167281516, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.762593269348145, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8627469539642334, "num_tokens": 341836061.0, "step": 8962 }, { "epoch": 1.1401857270067421, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.735397338867188, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8659740686416626, "num_tokens": 341875465.0, "step": 8963 }, { "epoch": 1.1403129372853327, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.720223426818848, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8610543012619019, "num_tokens": 341919795.0, "step": 8964 }, { "epoch": 1.1404401475639232, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.770777702331543, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8629564642906189, "num_tokens": 341953291.0, "step": 8965 }, { "epoch": 1.1405673578425137, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.74120044708252, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8849539160728455, "num_tokens": 341988364.0, "step": 8966 }, { "epoch": 1.1406945681211043, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75606632232666, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.87045818567276, "num_tokens": 342028478.0, "step": 8967 }, { "epoch": 1.1408217783996948, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.69563102722168, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8745021224021912, "num_tokens": 342067348.0, "step": 8968 }, { "epoch": 1.1409489886782853, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.795755386352539, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8627728223800659, "num_tokens": 342105845.0, "step": 8969 }, { "epoch": 1.1410761989568756, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.665678977966309, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8670846223831177, "num_tokens": 342142072.0, "step": 8970 }, { "epoch": 1.1412034092354661, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.739580154418945, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8701682090759277, "num_tokens": 342176592.0, "step": 8971 }, { "epoch": 1.1413306195140567, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.72628402709961, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.869803249835968, "num_tokens": 342211896.0, "step": 8972 }, { "epoch": 1.1414578297926472, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.705706596374512, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8699437975883484, "num_tokens": 342256078.0, "step": 8973 }, { "epoch": 1.1415850400712377, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.773920059204102, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8688517808914185, "num_tokens": 342296766.0, "step": 8974 }, { "epoch": 1.1417122503498283, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.719772338867188, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8665038347244263, "num_tokens": 342333689.0, "step": 8975 }, { "epoch": 1.1418394606284188, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.745019912719727, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8774429559707642, "num_tokens": 342369605.0, "step": 8976 }, { "epoch": 1.1419666709070093, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.713208198547363, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8761093020439148, "num_tokens": 342410806.0, "step": 8977 }, { "epoch": 1.1420938811855998, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.772826194763184, "learning_rate": 1e-06, "loss": 0.3346, "mean_token_accuracy": 0.8931068181991577, "num_tokens": 342446957.0, "step": 8978 }, { "epoch": 1.1422210914641904, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.755903244018555, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8561673164367676, "num_tokens": 342480717.0, "step": 8979 }, { "epoch": 1.142348301742781, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.82139778137207, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8746402263641357, "num_tokens": 342520431.0, "step": 8980 }, { "epoch": 1.1424755120213712, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.741482734680176, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.863551139831543, "num_tokens": 342561735.0, "step": 8981 }, { "epoch": 1.1426027222999617, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.793722152709961, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8783179521560669, "num_tokens": 342593899.0, "step": 8982 }, { "epoch": 1.1427299325785523, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75477123260498, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8654155731201172, "num_tokens": 342628044.0, "step": 8983 }, { "epoch": 1.1428571428571428, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.752140998840332, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8648133277893066, "num_tokens": 342665018.0, "step": 8984 }, { "epoch": 1.1429843531357333, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.7623929977417, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8562281727790833, "num_tokens": 342704832.0, "step": 8985 }, { "epoch": 1.1431115634143239, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.780355453491211, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8741464614868164, "num_tokens": 342741598.0, "step": 8986 }, { "epoch": 1.1432387736929144, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75391960144043, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8621629476547241, "num_tokens": 342777302.0, "step": 8987 }, { "epoch": 1.143365983971505, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.810687065124512, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8477915525436401, "num_tokens": 342809578.0, "step": 8988 }, { "epoch": 1.1434931942500954, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.73425006866455, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8682410717010498, "num_tokens": 342845809.0, "step": 8989 }, { "epoch": 1.143620404528686, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75225830078125, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8538965582847595, "num_tokens": 342886118.0, "step": 8990 }, { "epoch": 1.1437476148072765, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.788409233093262, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.863042414188385, "num_tokens": 342922517.0, "step": 8991 }, { "epoch": 1.143874825085867, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.765081405639648, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8603824377059937, "num_tokens": 342959741.0, "step": 8992 }, { "epoch": 1.1440020353644575, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.751978874206543, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8800250291824341, "num_tokens": 342991418.0, "step": 8993 }, { "epoch": 1.144129245643048, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.852496147155762, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8835830688476562, "num_tokens": 343024263.0, "step": 8994 }, { "epoch": 1.1442564559216384, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.73923397064209, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8680416345596313, "num_tokens": 343066388.0, "step": 8995 }, { "epoch": 1.144383666200229, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.77053451538086, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8680369257926941, "num_tokens": 343107631.0, "step": 8996 }, { "epoch": 1.1445108764788194, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.792855262756348, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8729443550109863, "num_tokens": 343143645.0, "step": 8997 }, { "epoch": 1.14463808675741, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.782493591308594, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8531649112701416, "num_tokens": 343182076.0, "step": 8998 }, { "epoch": 1.1447652970360005, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.711140632629395, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.885282576084137, "num_tokens": 343222808.0, "step": 8999 }, { "epoch": 1.144892507314591, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.754538536071777, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8634406328201294, "num_tokens": 343268229.0, "step": 9000 }, { "epoch": 1.1450197175931816, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.746709823608398, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8592831492424011, "num_tokens": 343304459.0, "step": 9001 }, { "epoch": 1.145146927871772, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.812597274780273, "learning_rate": 1e-06, "loss": 0.5392, "mean_token_accuracy": 0.8293066024780273, "num_tokens": 343343211.0, "step": 9002 }, { "epoch": 1.1452741381503626, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.672852516174316, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8627771139144897, "num_tokens": 343388498.0, "step": 9003 }, { "epoch": 1.1454013484289531, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.799063682556152, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.86503666639328, "num_tokens": 343430101.0, "step": 9004 }, { "epoch": 1.1455285587075437, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.78305721282959, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8546782732009888, "num_tokens": 343466454.0, "step": 9005 }, { "epoch": 1.145655768986134, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.711430549621582, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8815493583679199, "num_tokens": 343506960.0, "step": 9006 }, { "epoch": 1.1457829792647245, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.845179557800293, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8744242787361145, "num_tokens": 343541133.0, "step": 9007 }, { "epoch": 1.145910189543315, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.735872268676758, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.875327467918396, "num_tokens": 343580266.0, "step": 9008 }, { "epoch": 1.1460373998219056, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.738905906677246, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8640323281288147, "num_tokens": 343611691.0, "step": 9009 }, { "epoch": 1.146164610100496, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.738858222961426, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8722103834152222, "num_tokens": 343649355.0, "step": 9010 }, { "epoch": 1.1462918203790866, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.745280265808105, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8815375566482544, "num_tokens": 343681606.0, "step": 9011 }, { "epoch": 1.1464190306576771, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.779071807861328, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.877683162689209, "num_tokens": 343721592.0, "step": 9012 }, { "epoch": 1.1465462409362677, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.781050682067871, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8644352555274963, "num_tokens": 343764084.0, "step": 9013 }, { "epoch": 1.1466734512148582, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.740492820739746, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8688661456108093, "num_tokens": 343801213.0, "step": 9014 }, { "epoch": 1.1468006614934487, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.824274063110352, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8683758974075317, "num_tokens": 343839761.0, "step": 9015 }, { "epoch": 1.1469278717720393, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.780574798583984, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.86710524559021, "num_tokens": 343877792.0, "step": 9016 }, { "epoch": 1.1470550820506298, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.749393463134766, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8725268840789795, "num_tokens": 343915107.0, "step": 9017 }, { "epoch": 1.1471822923292203, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.792919158935547, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.871699333190918, "num_tokens": 343949783.0, "step": 9018 }, { "epoch": 1.1473095026078106, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.783596992492676, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8493062257766724, "num_tokens": 343991058.0, "step": 9019 }, { "epoch": 1.1474367128864011, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.796045303344727, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8671300411224365, "num_tokens": 344030424.0, "step": 9020 }, { "epoch": 1.1475639231649917, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.768791198730469, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8690529465675354, "num_tokens": 344071385.0, "step": 9021 }, { "epoch": 1.1476911334435822, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.786803245544434, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8762944340705872, "num_tokens": 344109725.0, "step": 9022 }, { "epoch": 1.1478183437221727, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.838221549987793, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8747860193252563, "num_tokens": 344143947.0, "step": 9023 }, { "epoch": 1.1479455540007633, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.815807342529297, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8568167686462402, "num_tokens": 344179969.0, "step": 9024 }, { "epoch": 1.1480727642793538, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.822787284851074, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8787597417831421, "num_tokens": 344212279.0, "step": 9025 }, { "epoch": 1.1481999745579443, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75993824005127, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8732014298439026, "num_tokens": 344248402.0, "step": 9026 }, { "epoch": 1.1483271848365348, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.826231956481934, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8647413849830627, "num_tokens": 344286557.0, "step": 9027 }, { "epoch": 1.1484543951151254, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.748696327209473, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8679865598678589, "num_tokens": 344331014.0, "step": 9028 }, { "epoch": 1.148581605393716, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.791091918945312, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8647421002388, "num_tokens": 344373239.0, "step": 9029 }, { "epoch": 1.1487088156723062, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.766705513000488, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8815417885780334, "num_tokens": 344409871.0, "step": 9030 }, { "epoch": 1.1488360259508967, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.804211616516113, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.856825590133667, "num_tokens": 344452108.0, "step": 9031 }, { "epoch": 1.1489632362294873, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.832978248596191, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8607810139656067, "num_tokens": 344489798.0, "step": 9032 }, { "epoch": 1.1490904465080778, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.760601997375488, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8752124309539795, "num_tokens": 344528194.0, "step": 9033 }, { "epoch": 1.1492176567866683, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.95152759552002, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.853083610534668, "num_tokens": 344559699.0, "step": 9034 }, { "epoch": 1.1493448670652588, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.809679985046387, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.860305666923523, "num_tokens": 344598234.0, "step": 9035 }, { "epoch": 1.1494720773438494, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.812593460083008, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8614395260810852, "num_tokens": 344634780.0, "step": 9036 }, { "epoch": 1.14959928762244, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.822836875915527, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8710876107215881, "num_tokens": 344677167.0, "step": 9037 }, { "epoch": 1.1497264979010304, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.783515930175781, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8550939559936523, "num_tokens": 344713824.0, "step": 9038 }, { "epoch": 1.149853708179621, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.866416931152344, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8545324802398682, "num_tokens": 344755814.0, "step": 9039 }, { "epoch": 1.1499809184582115, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.784586906433105, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.881675124168396, "num_tokens": 344795022.0, "step": 9040 }, { "epoch": 1.150108128736802, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.856258392333984, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8760209083557129, "num_tokens": 344833906.0, "step": 9041 }, { "epoch": 1.1502353390153925, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.775333404541016, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8716034889221191, "num_tokens": 344875975.0, "step": 9042 }, { "epoch": 1.150362549293983, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.834020614624023, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8711010217666626, "num_tokens": 344915433.0, "step": 9043 }, { "epoch": 1.1504897595725734, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.73013687133789, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8719879388809204, "num_tokens": 344949038.0, "step": 9044 }, { "epoch": 1.150616969851164, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.86546802520752, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8679838180541992, "num_tokens": 344988260.0, "step": 9045 }, { "epoch": 1.1507441801297544, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.706954956054688, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8800125122070312, "num_tokens": 345024355.0, "step": 9046 }, { "epoch": 1.150871390408345, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.861132621765137, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8762520551681519, "num_tokens": 345063041.0, "step": 9047 }, { "epoch": 1.1509986006869355, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.776817321777344, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8868180513381958, "num_tokens": 345097813.0, "step": 9048 }, { "epoch": 1.151125810965526, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.807034492492676, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8632827401161194, "num_tokens": 345134534.0, "step": 9049 }, { "epoch": 1.1512530212441165, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.904789924621582, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8716022968292236, "num_tokens": 345171018.0, "step": 9050 }, { "epoch": 1.151380231522707, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 12.332924842834473, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8784255981445312, "num_tokens": 345208006.0, "step": 9051 }, { "epoch": 1.1515074418012976, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.838638305664062, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8860218524932861, "num_tokens": 345246616.0, "step": 9052 }, { "epoch": 1.1516346520798881, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.662895202636719, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8843238353729248, "num_tokens": 345288095.0, "step": 9053 }, { "epoch": 1.1517618623584787, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.825151443481445, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8537657856941223, "num_tokens": 345333291.0, "step": 9054 }, { "epoch": 1.151889072637069, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.646674156188965, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8660629987716675, "num_tokens": 345375093.0, "step": 9055 }, { "epoch": 1.1520162829156595, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.790045738220215, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8610624670982361, "num_tokens": 345415547.0, "step": 9056 }, { "epoch": 1.15214349319425, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.775644302368164, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8711066842079163, "num_tokens": 345457340.0, "step": 9057 }, { "epoch": 1.1522707034728406, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.739078521728516, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8634510040283203, "num_tokens": 345494711.0, "step": 9058 }, { "epoch": 1.152397913751431, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.786242485046387, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8599417209625244, "num_tokens": 345534249.0, "step": 9059 }, { "epoch": 1.1525251240300216, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.809882164001465, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8661267757415771, "num_tokens": 345570832.0, "step": 9060 }, { "epoch": 1.1526523343086121, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.791274070739746, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8623813390731812, "num_tokens": 345609007.0, "step": 9061 }, { "epoch": 1.1527795445872027, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.719118118286133, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8788906335830688, "num_tokens": 345647979.0, "step": 9062 }, { "epoch": 1.1529067548657932, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.723907470703125, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8600532412528992, "num_tokens": 345681821.0, "step": 9063 }, { "epoch": 1.1530339651443837, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.79149341583252, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8576099872589111, "num_tokens": 345724274.0, "step": 9064 }, { "epoch": 1.1531611754229742, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.758028984069824, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.884047269821167, "num_tokens": 345763558.0, "step": 9065 }, { "epoch": 1.1532883857015648, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.763467788696289, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8619630336761475, "num_tokens": 345798486.0, "step": 9066 }, { "epoch": 1.1534155959801553, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.63796329498291, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8636977672576904, "num_tokens": 345839379.0, "step": 9067 }, { "epoch": 1.1535428062587456, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.79637336730957, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8636436462402344, "num_tokens": 345876675.0, "step": 9068 }, { "epoch": 1.1536700165373361, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.73487377166748, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.885771632194519, "num_tokens": 345913482.0, "step": 9069 }, { "epoch": 1.1537972268159267, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.798160552978516, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8636388778686523, "num_tokens": 345947671.0, "step": 9070 }, { "epoch": 1.1539244370945172, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.823770523071289, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.870136559009552, "num_tokens": 345981520.0, "step": 9071 }, { "epoch": 1.1540516473731077, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.739950180053711, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8503484725952148, "num_tokens": 346020050.0, "step": 9072 }, { "epoch": 1.1541788576516983, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.760683059692383, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8590623140335083, "num_tokens": 346050000.0, "step": 9073 }, { "epoch": 1.1543060679302888, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.750636100769043, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8616153001785278, "num_tokens": 346085038.0, "step": 9074 }, { "epoch": 1.1544332782088793, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.758939743041992, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8774008750915527, "num_tokens": 346125871.0, "step": 9075 }, { "epoch": 1.1545604884874698, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.764726638793945, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8760884404182434, "num_tokens": 346160460.0, "step": 9076 }, { "epoch": 1.1546876987660604, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.761131286621094, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8578853607177734, "num_tokens": 346201621.0, "step": 9077 }, { "epoch": 1.154814909044651, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.784716606140137, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8694015741348267, "num_tokens": 346242680.0, "step": 9078 }, { "epoch": 1.1549421193232412, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.761577606201172, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8582783937454224, "num_tokens": 346275774.0, "step": 9079 }, { "epoch": 1.1550693296018317, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.801297187805176, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8693889379501343, "num_tokens": 346317648.0, "step": 9080 }, { "epoch": 1.1551965398804223, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.835057258605957, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8705042600631714, "num_tokens": 346357736.0, "step": 9081 }, { "epoch": 1.1553237501590128, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.788991928100586, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8654825091362, "num_tokens": 346397961.0, "step": 9082 }, { "epoch": 1.1554509604376033, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.812897682189941, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8788750767707825, "num_tokens": 346438268.0, "step": 9083 }, { "epoch": 1.1555781707161938, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.756105422973633, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8585649728775024, "num_tokens": 346479024.0, "step": 9084 }, { "epoch": 1.1557053809947844, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.843367576599121, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8745347261428833, "num_tokens": 346523526.0, "step": 9085 }, { "epoch": 1.155832591273375, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.73570442199707, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8760271072387695, "num_tokens": 346561985.0, "step": 9086 }, { "epoch": 1.1559598015519654, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.80184555053711, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8537507057189941, "num_tokens": 346605600.0, "step": 9087 }, { "epoch": 1.156087011830556, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.779038429260254, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8808867335319519, "num_tokens": 346640798.0, "step": 9088 }, { "epoch": 1.1562142221091465, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.78315258026123, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8614394664764404, "num_tokens": 346678483.0, "step": 9089 }, { "epoch": 1.156341432387737, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.839312553405762, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8842186331748962, "num_tokens": 346714604.0, "step": 9090 }, { "epoch": 1.1564686426663275, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.770405769348145, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8782320618629456, "num_tokens": 346747203.0, "step": 9091 }, { "epoch": 1.156595852944918, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.843741416931152, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8663097023963928, "num_tokens": 346790650.0, "step": 9092 }, { "epoch": 1.1567230632235084, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.751087188720703, "learning_rate": 1e-06, "loss": 0.3428, "mean_token_accuracy": 0.8908569812774658, "num_tokens": 346822872.0, "step": 9093 }, { "epoch": 1.156850273502099, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.791576385498047, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8743683099746704, "num_tokens": 346857999.0, "step": 9094 }, { "epoch": 1.1569774837806894, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.773283004760742, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8753233551979065, "num_tokens": 346891804.0, "step": 9095 }, { "epoch": 1.15710469405928, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.782520294189453, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8667646050453186, "num_tokens": 346931554.0, "step": 9096 }, { "epoch": 1.1572319043378705, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.764120101928711, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8544669151306152, "num_tokens": 346965214.0, "step": 9097 }, { "epoch": 1.157359114616461, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.792022705078125, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8789144158363342, "num_tokens": 347001219.0, "step": 9098 }, { "epoch": 1.1574863248950515, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.786272048950195, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.850705623626709, "num_tokens": 347043212.0, "step": 9099 }, { "epoch": 1.157613535173642, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.711365699768066, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8798794746398926, "num_tokens": 347082249.0, "step": 9100 }, { "epoch": 1.1577407454522326, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.822267532348633, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8652900457382202, "num_tokens": 347119286.0, "step": 9101 }, { "epoch": 1.1578679557308231, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.75016975402832, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8809787034988403, "num_tokens": 347155549.0, "step": 9102 }, { "epoch": 1.1579951660094137, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.770991325378418, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8703404664993286, "num_tokens": 347196846.0, "step": 9103 }, { "epoch": 1.158122376288004, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.803345680236816, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8815946578979492, "num_tokens": 347234028.0, "step": 9104 }, { "epoch": 1.1582495865665945, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.754636764526367, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8797729015350342, "num_tokens": 347276931.0, "step": 9105 }, { "epoch": 1.158376796845185, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.717913627624512, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8646998405456543, "num_tokens": 347317139.0, "step": 9106 }, { "epoch": 1.1585040071237755, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.810800552368164, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8706915974617004, "num_tokens": 347355718.0, "step": 9107 }, { "epoch": 1.158631217402366, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.734797477722168, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8587503433227539, "num_tokens": 347393403.0, "step": 9108 }, { "epoch": 1.1587584276809566, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.788652420043945, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.875675618648529, "num_tokens": 347433736.0, "step": 9109 }, { "epoch": 1.1588856379595471, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.722967147827148, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8698115348815918, "num_tokens": 347474574.0, "step": 9110 }, { "epoch": 1.1590128482381377, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.749462127685547, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8738155364990234, "num_tokens": 347518785.0, "step": 9111 }, { "epoch": 1.1591400585167282, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.799223899841309, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8556984663009644, "num_tokens": 347557518.0, "step": 9112 }, { "epoch": 1.1592672687953187, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.772212982177734, "learning_rate": 1e-06, "loss": 0.4971, "mean_token_accuracy": 0.838778018951416, "num_tokens": 347600094.0, "step": 9113 }, { "epoch": 1.1593944790739092, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.825539588928223, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8671683669090271, "num_tokens": 347633709.0, "step": 9114 }, { "epoch": 1.1595216893524998, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.760539054870605, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8751640319824219, "num_tokens": 347670027.0, "step": 9115 }, { "epoch": 1.1596488996310903, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.760701179504395, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8725727796554565, "num_tokens": 347707488.0, "step": 9116 }, { "epoch": 1.1597761099096806, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.752960205078125, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8629152178764343, "num_tokens": 347746030.0, "step": 9117 }, { "epoch": 1.1599033201882711, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.771154403686523, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.860255241394043, "num_tokens": 347783409.0, "step": 9118 }, { "epoch": 1.1600305304668617, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.796158790588379, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8752132654190063, "num_tokens": 347821335.0, "step": 9119 }, { "epoch": 1.1601577407454522, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.768560409545898, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8787771463394165, "num_tokens": 347857590.0, "step": 9120 }, { "epoch": 1.1602849510240427, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.8174409866333, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.879745602607727, "num_tokens": 347897000.0, "step": 9121 }, { "epoch": 1.1604121613026332, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.775052070617676, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.868411660194397, "num_tokens": 347928947.0, "step": 9122 }, { "epoch": 1.1605393715812238, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.821964263916016, "learning_rate": 1e-06, "loss": 0.5221, "mean_token_accuracy": 0.8385027050971985, "num_tokens": 347959294.0, "step": 9123 }, { "epoch": 1.1606665818598143, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.758819580078125, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8732331395149231, "num_tokens": 347993103.0, "step": 9124 }, { "epoch": 1.1607937921384048, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.72515869140625, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8803350925445557, "num_tokens": 348028412.0, "step": 9125 }, { "epoch": 1.1609210024169954, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.752983093261719, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8512773513793945, "num_tokens": 348071429.0, "step": 9126 }, { "epoch": 1.161048212695586, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.70542049407959, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8725419044494629, "num_tokens": 348113222.0, "step": 9127 }, { "epoch": 1.1611754229741762, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.748523712158203, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8672463297843933, "num_tokens": 348150188.0, "step": 9128 }, { "epoch": 1.1613026332527667, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.753169059753418, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.871975302696228, "num_tokens": 348188547.0, "step": 9129 }, { "epoch": 1.1614298435313573, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.725765228271484, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8637635111808777, "num_tokens": 348228624.0, "step": 9130 }, { "epoch": 1.1615570538099478, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.736763954162598, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8644108176231384, "num_tokens": 348264932.0, "step": 9131 }, { "epoch": 1.1616842640885383, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.799344062805176, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8701934814453125, "num_tokens": 348307680.0, "step": 9132 }, { "epoch": 1.1618114743671288, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.757644653320312, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8826941847801208, "num_tokens": 348346449.0, "step": 9133 }, { "epoch": 1.1619386846457194, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.763579368591309, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8788198828697205, "num_tokens": 348386393.0, "step": 9134 }, { "epoch": 1.16206589492431, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.77050495147705, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8608355522155762, "num_tokens": 348429458.0, "step": 9135 }, { "epoch": 1.1621931052029004, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.866003036499023, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8556764125823975, "num_tokens": 348471365.0, "step": 9136 }, { "epoch": 1.162320315481491, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.705438613891602, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.853702962398529, "num_tokens": 348513425.0, "step": 9137 }, { "epoch": 1.1624475257600815, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.881088256835938, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8661743402481079, "num_tokens": 348551181.0, "step": 9138 }, { "epoch": 1.162574736038672, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.749030113220215, "learning_rate": 1e-06, "loss": 0.3344, "mean_token_accuracy": 0.8890529274940491, "num_tokens": 348583611.0, "step": 9139 }, { "epoch": 1.1627019463172625, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.826908111572266, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8659932613372803, "num_tokens": 348624281.0, "step": 9140 }, { "epoch": 1.162829156595853, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.86954402923584, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8667275309562683, "num_tokens": 348661416.0, "step": 9141 }, { "epoch": 1.1629563668744434, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.843332290649414, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8476405143737793, "num_tokens": 348703907.0, "step": 9142 }, { "epoch": 1.163083577153034, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.83558464050293, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8803172707557678, "num_tokens": 348735617.0, "step": 9143 }, { "epoch": 1.1632107874316244, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.819886207580566, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8659161925315857, "num_tokens": 348773552.0, "step": 9144 }, { "epoch": 1.163337997710215, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.781854629516602, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8692095875740051, "num_tokens": 348811274.0, "step": 9145 }, { "epoch": 1.1634652079888055, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.805490493774414, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8562376499176025, "num_tokens": 348847773.0, "step": 9146 }, { "epoch": 1.163592418267396, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.785006523132324, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8594139218330383, "num_tokens": 348886356.0, "step": 9147 }, { "epoch": 1.1637196285459865, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.846842765808105, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8788479566574097, "num_tokens": 348922892.0, "step": 9148 }, { "epoch": 1.163846838824577, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.76028823852539, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.8869006037712097, "num_tokens": 348954924.0, "step": 9149 }, { "epoch": 1.1639740491031676, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.819762229919434, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8741694688796997, "num_tokens": 348997335.0, "step": 9150 }, { "epoch": 1.1641012593817581, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.74365234375, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8573431968688965, "num_tokens": 349038221.0, "step": 9151 }, { "epoch": 1.1642284696603487, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.828673362731934, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.864365816116333, "num_tokens": 349076559.0, "step": 9152 }, { "epoch": 1.164355679938939, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.844675064086914, "learning_rate": 1e-06, "loss": 0.4783, "mean_token_accuracy": 0.8461951017379761, "num_tokens": 349112264.0, "step": 9153 }, { "epoch": 1.1644828902175295, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.83669662475586, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8804090023040771, "num_tokens": 349146571.0, "step": 9154 }, { "epoch": 1.16461010049612, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.765488624572754, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8755735158920288, "num_tokens": 349183674.0, "step": 9155 }, { "epoch": 1.1647373107747105, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.801259994506836, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8722236156463623, "num_tokens": 349227042.0, "step": 9156 }, { "epoch": 1.164864521053301, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.79572582244873, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8633085489273071, "num_tokens": 349264072.0, "step": 9157 }, { "epoch": 1.1649917313318916, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.821168899536133, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.869379460811615, "num_tokens": 349304427.0, "step": 9158 }, { "epoch": 1.1651189416104821, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.781535148620605, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.867565393447876, "num_tokens": 349346708.0, "step": 9159 }, { "epoch": 1.1652461518890727, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.800814628601074, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8667736053466797, "num_tokens": 349387716.0, "step": 9160 }, { "epoch": 1.1653733621676632, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.777785301208496, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.864924967288971, "num_tokens": 349425757.0, "step": 9161 }, { "epoch": 1.1655005724462537, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.813908576965332, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8748936057090759, "num_tokens": 349472097.0, "step": 9162 }, { "epoch": 1.1656277827248442, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.858771324157715, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8655952215194702, "num_tokens": 349513809.0, "step": 9163 }, { "epoch": 1.1657549930034348, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.825580596923828, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8847141265869141, "num_tokens": 349544906.0, "step": 9164 }, { "epoch": 1.1658822032820253, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.819923400878906, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8635271787643433, "num_tokens": 349582529.0, "step": 9165 }, { "epoch": 1.1660094135606156, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.796238899230957, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8624105453491211, "num_tokens": 349624871.0, "step": 9166 }, { "epoch": 1.1661366238392061, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.80544662475586, "learning_rate": 1e-06, "loss": 0.5113, "mean_token_accuracy": 0.8332291841506958, "num_tokens": 349660783.0, "step": 9167 }, { "epoch": 1.1662638341177967, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.804697036743164, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8786771297454834, "num_tokens": 349698430.0, "step": 9168 }, { "epoch": 1.1663910443963872, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.78563117980957, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8605951070785522, "num_tokens": 349734981.0, "step": 9169 }, { "epoch": 1.1665182546749777, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.84742546081543, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8627327084541321, "num_tokens": 349774267.0, "step": 9170 }, { "epoch": 1.1666454649535682, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.845741271972656, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8650574684143066, "num_tokens": 349811400.0, "step": 9171 }, { "epoch": 1.1667726752321588, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.79953384399414, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8596974015235901, "num_tokens": 349850031.0, "step": 9172 }, { "epoch": 1.1668998855107493, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.83362102508545, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8705389499664307, "num_tokens": 349889967.0, "step": 9173 }, { "epoch": 1.1670270957893398, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.877436637878418, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8622841238975525, "num_tokens": 349930276.0, "step": 9174 }, { "epoch": 1.1671543060679304, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.77658748626709, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8631671667098999, "num_tokens": 349969889.0, "step": 9175 }, { "epoch": 1.1672815163465209, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.856765747070312, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8646422624588013, "num_tokens": 350005729.0, "step": 9176 }, { "epoch": 1.1674087266251112, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.856951713562012, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8751876354217529, "num_tokens": 350043130.0, "step": 9177 }, { "epoch": 1.1675359369037017, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.791630744934082, "learning_rate": 1e-06, "loss": 0.353, "mean_token_accuracy": 0.8859524726867676, "num_tokens": 350081233.0, "step": 9178 }, { "epoch": 1.1676631471822922, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.785700798034668, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8544527888298035, "num_tokens": 350119403.0, "step": 9179 }, { "epoch": 1.1677903574608828, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.874549865722656, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8560389280319214, "num_tokens": 350159099.0, "step": 9180 }, { "epoch": 1.1679175677394733, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.812519073486328, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8682540655136108, "num_tokens": 350195642.0, "step": 9181 }, { "epoch": 1.1680447780180638, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.813758850097656, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8752107620239258, "num_tokens": 350236721.0, "step": 9182 }, { "epoch": 1.1681719882966544, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.828705787658691, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8643523454666138, "num_tokens": 350272691.0, "step": 9183 }, { "epoch": 1.168299198575245, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.777786254882812, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8664783239364624, "num_tokens": 350318299.0, "step": 9184 }, { "epoch": 1.1684264088538354, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.830195426940918, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8673824071884155, "num_tokens": 350360973.0, "step": 9185 }, { "epoch": 1.168553619132426, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.815563201904297, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8761399984359741, "num_tokens": 350400516.0, "step": 9186 }, { "epoch": 1.1686808294110165, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.893662452697754, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8633095026016235, "num_tokens": 350438147.0, "step": 9187 }, { "epoch": 1.168808039689607, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.759662628173828, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8731587529182434, "num_tokens": 350481604.0, "step": 9188 }, { "epoch": 1.1689352499681975, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.771879196166992, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8879823088645935, "num_tokens": 350518225.0, "step": 9189 }, { "epoch": 1.169062460246788, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.778355598449707, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8669246435165405, "num_tokens": 350553752.0, "step": 9190 }, { "epoch": 1.1691896705253784, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.828961372375488, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8729120492935181, "num_tokens": 350593676.0, "step": 9191 }, { "epoch": 1.169316880803969, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.765312194824219, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8662285804748535, "num_tokens": 350633618.0, "step": 9192 }, { "epoch": 1.1694440910825594, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.823084831237793, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8653784990310669, "num_tokens": 350678221.0, "step": 9193 }, { "epoch": 1.16957130136115, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.752456665039062, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8782246112823486, "num_tokens": 350718258.0, "step": 9194 }, { "epoch": 1.1696985116397405, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.831474304199219, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.878753125667572, "num_tokens": 350757370.0, "step": 9195 }, { "epoch": 1.169825721918331, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.802287101745605, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8686646223068237, "num_tokens": 350801394.0, "step": 9196 }, { "epoch": 1.1699529321969215, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.78377628326416, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8602755069732666, "num_tokens": 350840635.0, "step": 9197 }, { "epoch": 1.170080142475512, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.791165351867676, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.88390052318573, "num_tokens": 350879616.0, "step": 9198 }, { "epoch": 1.1702073527541026, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.788469314575195, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8676878213882446, "num_tokens": 350921710.0, "step": 9199 }, { "epoch": 1.1703345630326931, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.810583114624023, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8558723330497742, "num_tokens": 350964218.0, "step": 9200 }, { "epoch": 1.1704617733112836, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.813559532165527, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8775420188903809, "num_tokens": 351004511.0, "step": 9201 }, { "epoch": 1.170588983589874, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.840988159179688, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8632320165634155, "num_tokens": 351044583.0, "step": 9202 }, { "epoch": 1.1707161938684645, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.783581733703613, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8507639765739441, "num_tokens": 351083613.0, "step": 9203 }, { "epoch": 1.170843404147055, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.885326385498047, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.860285758972168, "num_tokens": 351117412.0, "step": 9204 }, { "epoch": 1.1709706144256455, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.798004150390625, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.854072093963623, "num_tokens": 351154779.0, "step": 9205 }, { "epoch": 1.171097824704236, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.811727523803711, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8785845041275024, "num_tokens": 351191412.0, "step": 9206 }, { "epoch": 1.1712250349828266, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.80574893951416, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8709957003593445, "num_tokens": 351235465.0, "step": 9207 }, { "epoch": 1.1713522452614171, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.784112930297852, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8736356496810913, "num_tokens": 351279326.0, "step": 9208 }, { "epoch": 1.1714794555400077, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.780500411987305, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8850077390670776, "num_tokens": 351316678.0, "step": 9209 }, { "epoch": 1.1716066658185982, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.858664512634277, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8641148805618286, "num_tokens": 351352641.0, "step": 9210 }, { "epoch": 1.1717338760971887, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.88895034790039, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8588129281997681, "num_tokens": 351385845.0, "step": 9211 }, { "epoch": 1.1718610863757792, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.849800109863281, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8744611144065857, "num_tokens": 351425220.0, "step": 9212 }, { "epoch": 1.1719882966543698, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.85510540008545, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8630104064941406, "num_tokens": 351467515.0, "step": 9213 }, { "epoch": 1.1721155069329603, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.809926986694336, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8660886287689209, "num_tokens": 351507571.0, "step": 9214 }, { "epoch": 1.1722427172115506, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.923542976379395, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8646805286407471, "num_tokens": 351546150.0, "step": 9215 }, { "epoch": 1.1723699274901411, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.8119478225708, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8613724708557129, "num_tokens": 351588574.0, "step": 9216 }, { "epoch": 1.1724971377687317, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.859682083129883, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8692305088043213, "num_tokens": 351626914.0, "step": 9217 }, { "epoch": 1.1726243480473222, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.864368438720703, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8578975796699524, "num_tokens": 351663574.0, "step": 9218 }, { "epoch": 1.1727515583259127, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.865242958068848, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8507221341133118, "num_tokens": 351703275.0, "step": 9219 }, { "epoch": 1.1728787686045032, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.79241943359375, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8576091527938843, "num_tokens": 351745713.0, "step": 9220 }, { "epoch": 1.1730059788830938, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.838869094848633, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8535009622573853, "num_tokens": 351780789.0, "step": 9221 }, { "epoch": 1.1731331891616843, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.901782989501953, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8779757618904114, "num_tokens": 351820352.0, "step": 9222 }, { "epoch": 1.1732603994402748, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.785425186157227, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8753801584243774, "num_tokens": 351860011.0, "step": 9223 }, { "epoch": 1.1733876097188654, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.846234321594238, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8620766997337341, "num_tokens": 351898742.0, "step": 9224 }, { "epoch": 1.1735148199974559, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.795970916748047, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8551361560821533, "num_tokens": 351940201.0, "step": 9225 }, { "epoch": 1.1736420302760462, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.854626655578613, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8659610748291016, "num_tokens": 351971041.0, "step": 9226 }, { "epoch": 1.1737692405546367, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.867212295532227, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8506857752799988, "num_tokens": 352013002.0, "step": 9227 }, { "epoch": 1.1738964508332272, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.82373046875, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8661664128303528, "num_tokens": 352051435.0, "step": 9228 }, { "epoch": 1.1740236611118178, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.807560920715332, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8735635876655579, "num_tokens": 352093295.0, "step": 9229 }, { "epoch": 1.1741508713904083, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.8304443359375, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8637707829475403, "num_tokens": 352136401.0, "step": 9230 }, { "epoch": 1.1742780816689988, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.760209083557129, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8783446550369263, "num_tokens": 352178696.0, "step": 9231 }, { "epoch": 1.1744052919475894, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.831313133239746, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8486552238464355, "num_tokens": 352214759.0, "step": 9232 }, { "epoch": 1.1745325022261799, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.797282218933105, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8808592557907104, "num_tokens": 352250520.0, "step": 9233 }, { "epoch": 1.1746597125047704, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.794486999511719, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8661655187606812, "num_tokens": 352288773.0, "step": 9234 }, { "epoch": 1.174786922783361, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.734542846679688, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8746718168258667, "num_tokens": 352326548.0, "step": 9235 }, { "epoch": 1.1749141330619515, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.777260780334473, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.876771092414856, "num_tokens": 352366537.0, "step": 9236 }, { "epoch": 1.175041343340542, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.779525756835938, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8511778712272644, "num_tokens": 352402390.0, "step": 9237 }, { "epoch": 1.1751685536191325, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.772052764892578, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8657369613647461, "num_tokens": 352442357.0, "step": 9238 }, { "epoch": 1.175295763897723, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.771454811096191, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8650951385498047, "num_tokens": 352476884.0, "step": 9239 }, { "epoch": 1.1754229741763134, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.784228324890137, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8679491877555847, "num_tokens": 352513448.0, "step": 9240 }, { "epoch": 1.175550184454904, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.817712783813477, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8863104581832886, "num_tokens": 352548420.0, "step": 9241 }, { "epoch": 1.1756773947334944, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.772906303405762, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8703203797340393, "num_tokens": 352592157.0, "step": 9242 }, { "epoch": 1.175804605012085, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.842652320861816, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8585302829742432, "num_tokens": 352626427.0, "step": 9243 }, { "epoch": 1.1759318152906755, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.82850456237793, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8762766718864441, "num_tokens": 352667075.0, "step": 9244 }, { "epoch": 1.176059025569266, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.75528621673584, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8648399114608765, "num_tokens": 352707114.0, "step": 9245 }, { "epoch": 1.1761862358478565, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.889078140258789, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8724627494812012, "num_tokens": 352738064.0, "step": 9246 }, { "epoch": 1.176313446126447, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.716368675231934, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8520556092262268, "num_tokens": 352779964.0, "step": 9247 }, { "epoch": 1.1764406564050376, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.848846435546875, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8753098249435425, "num_tokens": 352819574.0, "step": 9248 }, { "epoch": 1.1765678666836281, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.789891242980957, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8760843276977539, "num_tokens": 352859151.0, "step": 9249 }, { "epoch": 1.1766950769622184, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.874396324157715, "learning_rate": 1e-06, "loss": 0.5061, "mean_token_accuracy": 0.846560001373291, "num_tokens": 352901119.0, "step": 9250 }, { "epoch": 1.176822287240809, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.717533111572266, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8668544292449951, "num_tokens": 352945878.0, "step": 9251 }, { "epoch": 1.1769494975193995, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.83053207397461, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8614426851272583, "num_tokens": 352979819.0, "step": 9252 }, { "epoch": 1.17707670779799, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.800609588623047, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8570573329925537, "num_tokens": 353023552.0, "step": 9253 }, { "epoch": 1.1772039180765805, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.815535545349121, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8682858943939209, "num_tokens": 353058178.0, "step": 9254 }, { "epoch": 1.177331128355171, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.827890396118164, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8766007423400879, "num_tokens": 353091428.0, "step": 9255 }, { "epoch": 1.1774583386337616, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.861612319946289, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8703008890151978, "num_tokens": 353131943.0, "step": 9256 }, { "epoch": 1.1775855489123521, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.82364559173584, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.860480546951294, "num_tokens": 353171213.0, "step": 9257 }, { "epoch": 1.1777127591909426, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.831199645996094, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8755843639373779, "num_tokens": 353205057.0, "step": 9258 }, { "epoch": 1.1778399694695332, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.849507331848145, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8530724048614502, "num_tokens": 353251164.0, "step": 9259 }, { "epoch": 1.1779671797481237, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.78050422668457, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8720550537109375, "num_tokens": 353288202.0, "step": 9260 }, { "epoch": 1.1780943900267142, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.8164644241333, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8668166995048523, "num_tokens": 353327396.0, "step": 9261 }, { "epoch": 1.1782216003053048, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.79683780670166, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8564608097076416, "num_tokens": 353366512.0, "step": 9262 }, { "epoch": 1.1783488105838953, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.85605525970459, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8596105575561523, "num_tokens": 353408546.0, "step": 9263 }, { "epoch": 1.1784760208624856, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.874382972717285, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8748346567153931, "num_tokens": 353444950.0, "step": 9264 }, { "epoch": 1.1786032311410761, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.819507598876953, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8747338652610779, "num_tokens": 353483306.0, "step": 9265 }, { "epoch": 1.1787304414196667, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.813182830810547, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8656315803527832, "num_tokens": 353523038.0, "step": 9266 }, { "epoch": 1.1788576516982572, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.846231460571289, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.877130925655365, "num_tokens": 353564266.0, "step": 9267 }, { "epoch": 1.1789848619768477, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.842032432556152, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8708294630050659, "num_tokens": 353603558.0, "step": 9268 }, { "epoch": 1.1791120722554382, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.816926002502441, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8747571110725403, "num_tokens": 353648097.0, "step": 9269 }, { "epoch": 1.1792392825340288, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.789202690124512, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8516439199447632, "num_tokens": 353688970.0, "step": 9270 }, { "epoch": 1.1793664928126193, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.925113677978516, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8569929599761963, "num_tokens": 353732505.0, "step": 9271 }, { "epoch": 1.1794937030912098, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.76362133026123, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8759287595748901, "num_tokens": 353767460.0, "step": 9272 }, { "epoch": 1.1796209133698004, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.890193939208984, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8764077425003052, "num_tokens": 353801951.0, "step": 9273 }, { "epoch": 1.1797481236483909, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.814255714416504, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8753556609153748, "num_tokens": 353836815.0, "step": 9274 }, { "epoch": 1.1798753339269812, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.879792213439941, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8587737083435059, "num_tokens": 353876296.0, "step": 9275 }, { "epoch": 1.1800025442055717, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.852644920349121, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8513714671134949, "num_tokens": 353911792.0, "step": 9276 }, { "epoch": 1.1801297544841622, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.890010833740234, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8561608195304871, "num_tokens": 353949532.0, "step": 9277 }, { "epoch": 1.1802569647627528, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.81127643585205, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.855084240436554, "num_tokens": 353988622.0, "step": 9278 }, { "epoch": 1.1803841750413433, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.861024856567383, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8709700107574463, "num_tokens": 354028136.0, "step": 9279 }, { "epoch": 1.1805113853199338, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.820596694946289, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.869491457939148, "num_tokens": 354065029.0, "step": 9280 }, { "epoch": 1.1806385955985244, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.831743240356445, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8779313564300537, "num_tokens": 354099279.0, "step": 9281 }, { "epoch": 1.1807658058771149, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.840448379516602, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8627984523773193, "num_tokens": 354138870.0, "step": 9282 }, { "epoch": 1.1808930161557054, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.835306167602539, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8548431396484375, "num_tokens": 354176498.0, "step": 9283 }, { "epoch": 1.181020226434296, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.895277976989746, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8481792211532593, "num_tokens": 354210063.0, "step": 9284 }, { "epoch": 1.1811474367128865, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.852438926696777, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8602497577667236, "num_tokens": 354248095.0, "step": 9285 }, { "epoch": 1.181274646991477, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.782328605651855, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8631116151809692, "num_tokens": 354286897.0, "step": 9286 }, { "epoch": 1.1814018572700675, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.841161727905273, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8746506571769714, "num_tokens": 354330685.0, "step": 9287 }, { "epoch": 1.181529067548658, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.762594223022461, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8581243753433228, "num_tokens": 354371120.0, "step": 9288 }, { "epoch": 1.1816562778272484, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.837665557861328, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8646935224533081, "num_tokens": 354410825.0, "step": 9289 }, { "epoch": 1.1817834881058389, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.8113374710083, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8635009527206421, "num_tokens": 354453623.0, "step": 9290 }, { "epoch": 1.1819106983844294, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.844255447387695, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8668786287307739, "num_tokens": 354491854.0, "step": 9291 }, { "epoch": 1.18203790866302, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.787088394165039, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8535544872283936, "num_tokens": 354530788.0, "step": 9292 }, { "epoch": 1.1821651189416105, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.798757553100586, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8860633969306946, "num_tokens": 354568576.0, "step": 9293 }, { "epoch": 1.182292329220201, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.726943016052246, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8673413991928101, "num_tokens": 354611307.0, "step": 9294 }, { "epoch": 1.1824195394987915, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.846332550048828, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.873387336730957, "num_tokens": 354648810.0, "step": 9295 }, { "epoch": 1.182546749777382, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.794803619384766, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8687728047370911, "num_tokens": 354688411.0, "step": 9296 }, { "epoch": 1.1826739600559726, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.805595397949219, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8648892045021057, "num_tokens": 354734700.0, "step": 9297 }, { "epoch": 1.1828011703345631, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.847908020019531, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8632664680480957, "num_tokens": 354765517.0, "step": 9298 }, { "epoch": 1.1829283806131534, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.8309326171875, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8715969920158386, "num_tokens": 354808563.0, "step": 9299 }, { "epoch": 1.183055590891744, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.839099884033203, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8501197099685669, "num_tokens": 354846618.0, "step": 9300 }, { "epoch": 1.1831828011703345, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.885679244995117, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8696014881134033, "num_tokens": 354881293.0, "step": 9301 }, { "epoch": 1.183310011448925, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.833194732666016, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.864776074886322, "num_tokens": 354925239.0, "step": 9302 }, { "epoch": 1.1834372217275155, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.891515731811523, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8534601926803589, "num_tokens": 354967836.0, "step": 9303 }, { "epoch": 1.183564432006106, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.87309455871582, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8512927293777466, "num_tokens": 355003386.0, "step": 9304 }, { "epoch": 1.1836916422846966, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.841972351074219, "learning_rate": 1e-06, "loss": 0.4819, "mean_token_accuracy": 0.8457529544830322, "num_tokens": 355043391.0, "step": 9305 }, { "epoch": 1.1838188525632871, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.908143043518066, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.868136465549469, "num_tokens": 355082820.0, "step": 9306 }, { "epoch": 1.1839460628418776, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.888727188110352, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8873425126075745, "num_tokens": 355117531.0, "step": 9307 }, { "epoch": 1.1840732731204682, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.923221588134766, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8555215001106262, "num_tokens": 355159485.0, "step": 9308 }, { "epoch": 1.1842004833990587, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.85076904296875, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8621976971626282, "num_tokens": 355206051.0, "step": 9309 }, { "epoch": 1.1843276936776492, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.939131736755371, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8583004474639893, "num_tokens": 355242933.0, "step": 9310 }, { "epoch": 1.1844549039562398, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.937743186950684, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8798856735229492, "num_tokens": 355286311.0, "step": 9311 }, { "epoch": 1.1845821142348303, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.916375160217285, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8670996427536011, "num_tokens": 355327844.0, "step": 9312 }, { "epoch": 1.1847093245134206, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.867409706115723, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8755951523780823, "num_tokens": 355368977.0, "step": 9313 }, { "epoch": 1.1848365347920111, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.84247875213623, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8570095300674438, "num_tokens": 355411479.0, "step": 9314 }, { "epoch": 1.1849637450706016, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.949711799621582, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8844823241233826, "num_tokens": 355447370.0, "step": 9315 }, { "epoch": 1.1850909553491922, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.830716133117676, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8637953996658325, "num_tokens": 355490335.0, "step": 9316 }, { "epoch": 1.1852181656277827, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.851250648498535, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8678522109985352, "num_tokens": 355527843.0, "step": 9317 }, { "epoch": 1.1853453759063732, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.906438827514648, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.86385178565979, "num_tokens": 355566094.0, "step": 9318 }, { "epoch": 1.1854725861849638, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.84442138671875, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8748770356178284, "num_tokens": 355609505.0, "step": 9319 }, { "epoch": 1.1855997964635543, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.879169464111328, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8423799276351929, "num_tokens": 355650930.0, "step": 9320 }, { "epoch": 1.1857270067421448, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.898285865783691, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8760558366775513, "num_tokens": 355684075.0, "step": 9321 }, { "epoch": 1.1858542170207353, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.806074142456055, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8718569278717041, "num_tokens": 355723567.0, "step": 9322 }, { "epoch": 1.1859814272993259, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.827437400817871, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8532208204269409, "num_tokens": 355764915.0, "step": 9323 }, { "epoch": 1.1861086375779162, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.840411186218262, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8685783743858337, "num_tokens": 355804950.0, "step": 9324 }, { "epoch": 1.1862358478565067, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.898487091064453, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8748164772987366, "num_tokens": 355844155.0, "step": 9325 }, { "epoch": 1.1863630581350972, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.828120231628418, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8827955722808838, "num_tokens": 355883971.0, "step": 9326 }, { "epoch": 1.1864902684136878, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.831076622009277, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8620061874389648, "num_tokens": 355922271.0, "step": 9327 }, { "epoch": 1.1866174786922783, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.860077857971191, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8656777143478394, "num_tokens": 355961087.0, "step": 9328 }, { "epoch": 1.1867446889708688, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.832561492919922, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8704403638839722, "num_tokens": 355999553.0, "step": 9329 }, { "epoch": 1.1868718992494594, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.895153045654297, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8715725541114807, "num_tokens": 356038503.0, "step": 9330 }, { "epoch": 1.1869991095280499, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.841110229492188, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8778781294822693, "num_tokens": 356078538.0, "step": 9331 }, { "epoch": 1.1871263198066404, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.851896286010742, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8790286183357239, "num_tokens": 356115031.0, "step": 9332 }, { "epoch": 1.187253530085231, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.841851234436035, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8807556629180908, "num_tokens": 356154751.0, "step": 9333 }, { "epoch": 1.1873807403638215, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.90682315826416, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8519421219825745, "num_tokens": 356192105.0, "step": 9334 }, { "epoch": 1.187507950642412, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.898968696594238, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8571373224258423, "num_tokens": 356224964.0, "step": 9335 }, { "epoch": 1.1876351609210025, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.835057258605957, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8715962767601013, "num_tokens": 356261565.0, "step": 9336 }, { "epoch": 1.187762371199593, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.808609008789062, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8679447174072266, "num_tokens": 356301886.0, "step": 9337 }, { "epoch": 1.1878895814781834, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.785165786743164, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8748010396957397, "num_tokens": 356339481.0, "step": 9338 }, { "epoch": 1.1880167917567739, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.784749031066895, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8437454700469971, "num_tokens": 356387038.0, "step": 9339 }, { "epoch": 1.1881440020353644, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.857930183410645, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8640846014022827, "num_tokens": 356433256.0, "step": 9340 }, { "epoch": 1.188271212313955, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.825972557067871, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8564602136611938, "num_tokens": 356471282.0, "step": 9341 }, { "epoch": 1.1883984225925455, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.832979202270508, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8731693029403687, "num_tokens": 356508951.0, "step": 9342 }, { "epoch": 1.188525632871136, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.830891609191895, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8693310022354126, "num_tokens": 356544487.0, "step": 9343 }, { "epoch": 1.1886528431497265, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.879639625549316, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8737897276878357, "num_tokens": 356580873.0, "step": 9344 }, { "epoch": 1.188780053428317, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.865145683288574, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8650794625282288, "num_tokens": 356613622.0, "step": 9345 }, { "epoch": 1.1889072637069076, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.878081321716309, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8441575765609741, "num_tokens": 356660058.0, "step": 9346 }, { "epoch": 1.189034473985498, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.841950416564941, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8665334582328796, "num_tokens": 356701614.0, "step": 9347 }, { "epoch": 1.1891616842640884, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.871285438537598, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8815341591835022, "num_tokens": 356737105.0, "step": 9348 }, { "epoch": 1.189288894542679, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.807077407836914, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8584787845611572, "num_tokens": 356779382.0, "step": 9349 }, { "epoch": 1.1894161048212695, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.854416847229004, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8620758652687073, "num_tokens": 356813836.0, "step": 9350 }, { "epoch": 1.18954331509986, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.875534057617188, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.857058584690094, "num_tokens": 356852591.0, "step": 9351 }, { "epoch": 1.1896705253784505, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.901241302490234, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8649924993515015, "num_tokens": 356889940.0, "step": 9352 }, { "epoch": 1.189797735657041, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.797330856323242, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8700424432754517, "num_tokens": 356927878.0, "step": 9353 }, { "epoch": 1.1899249459356316, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.894596099853516, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8763136267662048, "num_tokens": 356963339.0, "step": 9354 }, { "epoch": 1.1900521562142221, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.806092262268066, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8715870976448059, "num_tokens": 357002999.0, "step": 9355 }, { "epoch": 1.1901793664928126, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.871432304382324, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8701239228248596, "num_tokens": 357044305.0, "step": 9356 }, { "epoch": 1.1903065767714032, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.85340690612793, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8685989379882812, "num_tokens": 357082197.0, "step": 9357 }, { "epoch": 1.1904337870499937, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.895162582397461, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8493168950080872, "num_tokens": 357119541.0, "step": 9358 }, { "epoch": 1.1905609973285842, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.872637748718262, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8889175653457642, "num_tokens": 357159176.0, "step": 9359 }, { "epoch": 1.1906882076071748, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.883834838867188, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8665012121200562, "num_tokens": 357199204.0, "step": 9360 }, { "epoch": 1.1908154178857653, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.901684761047363, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8807417154312134, "num_tokens": 357236900.0, "step": 9361 }, { "epoch": 1.1909426281643556, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.863396644592285, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.867006778717041, "num_tokens": 357271622.0, "step": 9362 }, { "epoch": 1.1910698384429461, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.826889991760254, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.877699077129364, "num_tokens": 357311145.0, "step": 9363 }, { "epoch": 1.1911970487215366, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.936872482299805, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8841238617897034, "num_tokens": 357344609.0, "step": 9364 }, { "epoch": 1.1913242590001272, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.877654075622559, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8645403385162354, "num_tokens": 357378896.0, "step": 9365 }, { "epoch": 1.1914514692787177, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.804447174072266, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8708820939064026, "num_tokens": 357424730.0, "step": 9366 }, { "epoch": 1.1915786795573082, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.872261047363281, "learning_rate": 1e-06, "loss": 0.4816, "mean_token_accuracy": 0.8484113216400146, "num_tokens": 357461152.0, "step": 9367 }, { "epoch": 1.1917058898358988, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.900321006774902, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8606821298599243, "num_tokens": 357502305.0, "step": 9368 }, { "epoch": 1.1918331001144893, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.80688762664795, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8615477085113525, "num_tokens": 357539187.0, "step": 9369 }, { "epoch": 1.1919603103930798, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.902471542358398, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8570213317871094, "num_tokens": 357569520.0, "step": 9370 }, { "epoch": 1.1920875206716703, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.834895133972168, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8668491244316101, "num_tokens": 357605253.0, "step": 9371 }, { "epoch": 1.1922147309502609, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.915328025817871, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8782933950424194, "num_tokens": 357639701.0, "step": 9372 }, { "epoch": 1.1923419412288512, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.81898307800293, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8751081228256226, "num_tokens": 357675611.0, "step": 9373 }, { "epoch": 1.1924691515074417, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.865779876708984, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.879021406173706, "num_tokens": 357708521.0, "step": 9374 }, { "epoch": 1.1925963617860322, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.851040840148926, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8576712608337402, "num_tokens": 357749970.0, "step": 9375 }, { "epoch": 1.1927235720646228, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.974626541137695, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8773012161254883, "num_tokens": 357783287.0, "step": 9376 }, { "epoch": 1.1928507823432133, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.866684913635254, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8685652017593384, "num_tokens": 357818883.0, "step": 9377 }, { "epoch": 1.1929779926218038, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.900775909423828, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8801321983337402, "num_tokens": 357862862.0, "step": 9378 }, { "epoch": 1.1931052029003943, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.88818073272705, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8565137386322021, "num_tokens": 357901574.0, "step": 9379 }, { "epoch": 1.1932324131789849, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.946069717407227, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8655604124069214, "num_tokens": 357941822.0, "step": 9380 }, { "epoch": 1.1933596234575754, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.824386596679688, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.86016845703125, "num_tokens": 357982820.0, "step": 9381 }, { "epoch": 1.193486833736166, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.93537712097168, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8647847175598145, "num_tokens": 358020491.0, "step": 9382 }, { "epoch": 1.1936140440147565, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.868694305419922, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8692572116851807, "num_tokens": 358059775.0, "step": 9383 }, { "epoch": 1.193741254293347, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.841706275939941, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8743670582771301, "num_tokens": 358098672.0, "step": 9384 }, { "epoch": 1.1938684645719375, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.88224983215332, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8593592047691345, "num_tokens": 358140850.0, "step": 9385 }, { "epoch": 1.193995674850528, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.926316261291504, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8737280368804932, "num_tokens": 358175503.0, "step": 9386 }, { "epoch": 1.1941228851291183, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.788630485534668, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8706785440444946, "num_tokens": 358210615.0, "step": 9387 }, { "epoch": 1.1942500954077089, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.968062400817871, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8695060610771179, "num_tokens": 358246605.0, "step": 9388 }, { "epoch": 1.1943773056862994, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.856513023376465, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8637489080429077, "num_tokens": 358283136.0, "step": 9389 }, { "epoch": 1.19450451596489, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.887704849243164, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8755301833152771, "num_tokens": 358325437.0, "step": 9390 }, { "epoch": 1.1946317262434805, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.903202056884766, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8660624623298645, "num_tokens": 358365360.0, "step": 9391 }, { "epoch": 1.194758936522071, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.844715118408203, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8837652206420898, "num_tokens": 358406502.0, "step": 9392 }, { "epoch": 1.1948861468006615, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.950510025024414, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8623912334442139, "num_tokens": 358437220.0, "step": 9393 }, { "epoch": 1.195013357079252, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.904841423034668, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8558925986289978, "num_tokens": 358475314.0, "step": 9394 }, { "epoch": 1.1951405673578426, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.887392044067383, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8778899908065796, "num_tokens": 358520777.0, "step": 9395 }, { "epoch": 1.195267777636433, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.865493774414062, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8867945671081543, "num_tokens": 358557973.0, "step": 9396 }, { "epoch": 1.1953949879150234, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.934418678283691, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8745987415313721, "num_tokens": 358596795.0, "step": 9397 }, { "epoch": 1.195522198193614, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.896607398986816, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8677162528038025, "num_tokens": 358632146.0, "step": 9398 }, { "epoch": 1.1956494084722045, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.91647720336914, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8760846257209778, "num_tokens": 358672194.0, "step": 9399 }, { "epoch": 1.195776618750795, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.935790061950684, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8751950263977051, "num_tokens": 358710192.0, "step": 9400 }, { "epoch": 1.1959038290293855, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.94977855682373, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8545023202896118, "num_tokens": 358749284.0, "step": 9401 }, { "epoch": 1.196031039307976, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.909372329711914, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8636982440948486, "num_tokens": 358782830.0, "step": 9402 }, { "epoch": 1.1961582495865666, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.927132606506348, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8652991056442261, "num_tokens": 358820696.0, "step": 9403 }, { "epoch": 1.196285459865157, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.992128372192383, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8744752407073975, "num_tokens": 358853616.0, "step": 9404 }, { "epoch": 1.1964126701437476, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.895834922790527, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8758201599121094, "num_tokens": 358893420.0, "step": 9405 }, { "epoch": 1.1965398804223382, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.901296615600586, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8695822954177856, "num_tokens": 358930704.0, "step": 9406 }, { "epoch": 1.1966670907009287, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.910961151123047, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8807367086410522, "num_tokens": 358967730.0, "step": 9407 }, { "epoch": 1.1967943009795192, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.974327087402344, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.879921019077301, "num_tokens": 359007241.0, "step": 9408 }, { "epoch": 1.1969215112581097, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.828897476196289, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8558050394058228, "num_tokens": 359045573.0, "step": 9409 }, { "epoch": 1.1970487215367003, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.919628143310547, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.87138432264328, "num_tokens": 359083420.0, "step": 9410 }, { "epoch": 1.1971759318152906, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.85499095916748, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8809937238693237, "num_tokens": 359117555.0, "step": 9411 }, { "epoch": 1.1973031420938811, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.902512550354004, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8651124835014343, "num_tokens": 359154597.0, "step": 9412 }, { "epoch": 1.1974303523724716, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.835097312927246, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8625880479812622, "num_tokens": 359190074.0, "step": 9413 }, { "epoch": 1.1975575626510622, "ewc_loss": 0.02587890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.833841323852539, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8562600612640381, "num_tokens": 359232711.0, "step": 9414 }, { "epoch": 1.1976847729296527, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.93964672088623, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.868396520614624, "num_tokens": 359265582.0, "step": 9415 }, { "epoch": 1.1978119832082432, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.867156982421875, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8752856254577637, "num_tokens": 359303177.0, "step": 9416 }, { "epoch": 1.1979391934868338, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.838016510009766, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8775366544723511, "num_tokens": 359338712.0, "step": 9417 }, { "epoch": 1.1980664037654243, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.86553955078125, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8762830495834351, "num_tokens": 359372849.0, "step": 9418 }, { "epoch": 1.1981936140440148, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.955275535583496, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8729648590087891, "num_tokens": 359407981.0, "step": 9419 }, { "epoch": 1.1983208243226053, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.868796348571777, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8471434116363525, "num_tokens": 359450180.0, "step": 9420 }, { "epoch": 1.1984480346011959, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.892143249511719, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8677536249160767, "num_tokens": 359487909.0, "step": 9421 }, { "epoch": 1.1985752448797862, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.93398666381836, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8795361518859863, "num_tokens": 359522609.0, "step": 9422 }, { "epoch": 1.1987024551583767, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.882522583007812, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.864414632320404, "num_tokens": 359560735.0, "step": 9423 }, { "epoch": 1.1988296654369672, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.820844650268555, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8613208532333374, "num_tokens": 359602994.0, "step": 9424 }, { "epoch": 1.1989568757155578, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.951210975646973, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.867613673210144, "num_tokens": 359638864.0, "step": 9425 }, { "epoch": 1.1990840859941483, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.853013038635254, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8758482933044434, "num_tokens": 359674202.0, "step": 9426 }, { "epoch": 1.1992112962727388, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.949005126953125, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8731764554977417, "num_tokens": 359710408.0, "step": 9427 }, { "epoch": 1.1993385065513293, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.903569221496582, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8763144016265869, "num_tokens": 359749068.0, "step": 9428 }, { "epoch": 1.1994657168299199, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.789937019348145, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.868187427520752, "num_tokens": 359788449.0, "step": 9429 }, { "epoch": 1.1995929271085104, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.985957145690918, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8626114130020142, "num_tokens": 359824167.0, "step": 9430 }, { "epoch": 1.199720137387101, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.909625053405762, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8633818030357361, "num_tokens": 359861476.0, "step": 9431 }, { "epoch": 1.1998473476656915, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.896435737609863, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8661593198776245, "num_tokens": 359906463.0, "step": 9432 }, { "epoch": 1.199974557944282, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.930569648742676, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8716388940811157, "num_tokens": 359940576.0, "step": 9433 }, { "epoch": 1.2001017682228725, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.93454647064209, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8779072165489197, "num_tokens": 359973616.0, "step": 9434 }, { "epoch": 1.200228978501463, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.920388221740723, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.880792498588562, "num_tokens": 360011059.0, "step": 9435 }, { "epoch": 1.2003561887800533, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.878327369689941, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8778377771377563, "num_tokens": 360048307.0, "step": 9436 }, { "epoch": 1.2004833990586439, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.917215347290039, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8680592179298401, "num_tokens": 360086946.0, "step": 9437 }, { "epoch": 1.2006106093372344, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.865588188171387, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8732534050941467, "num_tokens": 360121908.0, "step": 9438 }, { "epoch": 1.200737819615825, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.899178504943848, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8588064908981323, "num_tokens": 360164349.0, "step": 9439 }, { "epoch": 1.2008650298944155, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.887130737304688, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8829284906387329, "num_tokens": 360197660.0, "step": 9440 }, { "epoch": 1.200992240173006, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.911126136779785, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8473418951034546, "num_tokens": 360240297.0, "step": 9441 }, { "epoch": 1.2011194504515965, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.824955940246582, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8577514886856079, "num_tokens": 360281566.0, "step": 9442 }, { "epoch": 1.201246660730187, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.978761672973633, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.874267041683197, "num_tokens": 360313927.0, "step": 9443 }, { "epoch": 1.2013738710087776, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.860282897949219, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8721601366996765, "num_tokens": 360349369.0, "step": 9444 }, { "epoch": 1.201501081287368, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.869444847106934, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8674795031547546, "num_tokens": 360386996.0, "step": 9445 }, { "epoch": 1.2016282915659584, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.898825645446777, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8723294734954834, "num_tokens": 360424883.0, "step": 9446 }, { "epoch": 1.201755501844549, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.90956974029541, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8733116984367371, "num_tokens": 360457992.0, "step": 9447 }, { "epoch": 1.2018827121231395, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.858855247497559, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8534113168716431, "num_tokens": 360497934.0, "step": 9448 }, { "epoch": 1.20200992240173, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.900518417358398, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8658641576766968, "num_tokens": 360533928.0, "step": 9449 }, { "epoch": 1.2021371326803205, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.927608489990234, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8709047436714172, "num_tokens": 360571076.0, "step": 9450 }, { "epoch": 1.202264342958911, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.848077774047852, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8604334592819214, "num_tokens": 360611957.0, "step": 9451 }, { "epoch": 1.2023915532375016, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.878222465515137, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8617905974388123, "num_tokens": 360647823.0, "step": 9452 }, { "epoch": 1.202518763516092, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.900559425354004, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8592803478240967, "num_tokens": 360688341.0, "step": 9453 }, { "epoch": 1.2026459737946826, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.839374542236328, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8605624437332153, "num_tokens": 360730417.0, "step": 9454 }, { "epoch": 1.2027731840732732, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.924334526062012, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8491013050079346, "num_tokens": 360771931.0, "step": 9455 }, { "epoch": 1.2029003943518637, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.860732078552246, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8658809661865234, "num_tokens": 360811460.0, "step": 9456 }, { "epoch": 1.2030276046304542, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.87657356262207, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8797721862792969, "num_tokens": 360849476.0, "step": 9457 }, { "epoch": 1.2031548149090447, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.871030807495117, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8579561710357666, "num_tokens": 360891178.0, "step": 9458 }, { "epoch": 1.2032820251876353, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.922073364257812, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8648536205291748, "num_tokens": 360930503.0, "step": 9459 }, { "epoch": 1.2034092354662256, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.915152549743652, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.864235520362854, "num_tokens": 360963730.0, "step": 9460 }, { "epoch": 1.203536445744816, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.83590316772461, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.85364830493927, "num_tokens": 361003796.0, "step": 9461 }, { "epoch": 1.2036636560234066, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.896724700927734, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8691009283065796, "num_tokens": 361044818.0, "step": 9462 }, { "epoch": 1.2037908663019972, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.82606315612793, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8785574436187744, "num_tokens": 361080692.0, "step": 9463 }, { "epoch": 1.2039180765805877, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.909918785095215, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8589948415756226, "num_tokens": 361116609.0, "step": 9464 }, { "epoch": 1.2040452868591782, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.893827438354492, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8616207242012024, "num_tokens": 361154559.0, "step": 9465 }, { "epoch": 1.2041724971377687, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.869356155395508, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8761675357818604, "num_tokens": 361191005.0, "step": 9466 }, { "epoch": 1.2042997074163593, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.880334854125977, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8647603392601013, "num_tokens": 361230191.0, "step": 9467 }, { "epoch": 1.2044269176949498, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.81406021118164, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8615808486938477, "num_tokens": 361269132.0, "step": 9468 }, { "epoch": 1.2045541279735403, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.868916511535645, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8603996634483337, "num_tokens": 361306674.0, "step": 9469 }, { "epoch": 1.2046813382521309, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.915021896362305, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.869938850402832, "num_tokens": 361348309.0, "step": 9470 }, { "epoch": 1.2048085485307212, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.794073104858398, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.863447904586792, "num_tokens": 361386824.0, "step": 9471 }, { "epoch": 1.2049357588093117, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.878738403320312, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8498064279556274, "num_tokens": 361426590.0, "step": 9472 }, { "epoch": 1.2050629690879022, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.910479545593262, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8538199663162231, "num_tokens": 361463259.0, "step": 9473 }, { "epoch": 1.2051901793664928, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.874980926513672, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8770748376846313, "num_tokens": 361498750.0, "step": 9474 }, { "epoch": 1.2053173896450833, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.854253768920898, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8575522303581238, "num_tokens": 361537101.0, "step": 9475 }, { "epoch": 1.2054445999236738, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.906699180603027, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8704173564910889, "num_tokens": 361572130.0, "step": 9476 }, { "epoch": 1.2055718102022643, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.831157684326172, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8838146924972534, "num_tokens": 361608484.0, "step": 9477 }, { "epoch": 1.2056990204808549, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.935233116149902, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.8848018646240234, "num_tokens": 361650361.0, "step": 9478 }, { "epoch": 1.2058262307594454, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.888130187988281, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8755607604980469, "num_tokens": 361680670.0, "step": 9479 }, { "epoch": 1.205953441038036, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.874980926513672, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8731845617294312, "num_tokens": 361723504.0, "step": 9480 }, { "epoch": 1.2060806513166265, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.887189865112305, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8738112449645996, "num_tokens": 361765850.0, "step": 9481 }, { "epoch": 1.206207861595217, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.939952850341797, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8588000535964966, "num_tokens": 361804375.0, "step": 9482 }, { "epoch": 1.2063350718738075, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.988006591796875, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8649393320083618, "num_tokens": 361834714.0, "step": 9483 }, { "epoch": 1.206462282152398, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.858481407165527, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8677833080291748, "num_tokens": 361877568.0, "step": 9484 }, { "epoch": 1.2065894924309883, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.899614334106445, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8636075258255005, "num_tokens": 361914785.0, "step": 9485 }, { "epoch": 1.2067167027095789, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.9298095703125, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8751919269561768, "num_tokens": 361956781.0, "step": 9486 }, { "epoch": 1.2068439129881694, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.798370361328125, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8770911693572998, "num_tokens": 361997408.0, "step": 9487 }, { "epoch": 1.20697112326676, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.943401336669922, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8695721626281738, "num_tokens": 362030226.0, "step": 9488 }, { "epoch": 1.2070983335453505, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.773744583129883, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8781891465187073, "num_tokens": 362068370.0, "step": 9489 }, { "epoch": 1.207225543823941, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.857222557067871, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8760938048362732, "num_tokens": 362108569.0, "step": 9490 }, { "epoch": 1.2073527541025315, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.937204360961914, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8602044582366943, "num_tokens": 362140476.0, "step": 9491 }, { "epoch": 1.207479964381122, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.927621841430664, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8759052753448486, "num_tokens": 362171132.0, "step": 9492 }, { "epoch": 1.2076071746597126, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.909017562866211, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8649340867996216, "num_tokens": 362204791.0, "step": 9493 }, { "epoch": 1.207734384938303, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.87343978881836, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8708127737045288, "num_tokens": 362242360.0, "step": 9494 }, { "epoch": 1.2078615952168934, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.970264434814453, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8697145581245422, "num_tokens": 362276558.0, "step": 9495 }, { "epoch": 1.207988805495484, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.970789909362793, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8541587591171265, "num_tokens": 362307055.0, "step": 9496 }, { "epoch": 1.2081160157740745, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.894255638122559, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8794567584991455, "num_tokens": 362342433.0, "step": 9497 }, { "epoch": 1.208243226052665, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.884268760681152, "learning_rate": 1e-06, "loss": 0.3359, "mean_token_accuracy": 0.889031171798706, "num_tokens": 362377008.0, "step": 9498 }, { "epoch": 1.2083704363312555, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.873234748840332, "learning_rate": 1e-06, "loss": 0.3486, "mean_token_accuracy": 0.8823855519294739, "num_tokens": 362412089.0, "step": 9499 }, { "epoch": 1.208497646609846, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.924860954284668, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.859250545501709, "num_tokens": 362455817.0, "step": 9500 }, { "epoch": 1.2086248568884366, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.897115707397461, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8677147626876831, "num_tokens": 362498859.0, "step": 9501 }, { "epoch": 1.208752067167027, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.876998901367188, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8670496940612793, "num_tokens": 362542346.0, "step": 9502 }, { "epoch": 1.2088792774456176, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.932793617248535, "learning_rate": 1e-06, "loss": 0.3379, "mean_token_accuracy": 0.886926531791687, "num_tokens": 362575360.0, "step": 9503 }, { "epoch": 1.2090064877242082, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.906938552856445, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.868332028388977, "num_tokens": 362614737.0, "step": 9504 }, { "epoch": 1.2091336980027987, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.887453079223633, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8626941442489624, "num_tokens": 362653756.0, "step": 9505 }, { "epoch": 1.2092609082813892, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.909793853759766, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8702393770217896, "num_tokens": 362690745.0, "step": 9506 }, { "epoch": 1.2093881185599797, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.932517051696777, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8772013187408447, "num_tokens": 362726328.0, "step": 9507 }, { "epoch": 1.2095153288385703, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.88143539428711, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.87092125415802, "num_tokens": 362764285.0, "step": 9508 }, { "epoch": 1.2096425391171606, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.872888565063477, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8507034182548523, "num_tokens": 362794697.0, "step": 9509 }, { "epoch": 1.209769749395751, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.950160026550293, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8673646450042725, "num_tokens": 362831618.0, "step": 9510 }, { "epoch": 1.2098969596743416, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.886812210083008, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8764406442642212, "num_tokens": 362866404.0, "step": 9511 }, { "epoch": 1.2100241699529322, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.917219161987305, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8845874071121216, "num_tokens": 362902554.0, "step": 9512 }, { "epoch": 1.2101513802315227, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.871159553527832, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8624521493911743, "num_tokens": 362944169.0, "step": 9513 }, { "epoch": 1.2102785905101132, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.929925918579102, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8473982810974121, "num_tokens": 362983789.0, "step": 9514 }, { "epoch": 1.2104058007887037, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.871560096740723, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8742872476577759, "num_tokens": 363021551.0, "step": 9515 }, { "epoch": 1.2105330110672943, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.90121841430664, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8518555164337158, "num_tokens": 363064951.0, "step": 9516 }, { "epoch": 1.2106602213458848, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.870058059692383, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8679618239402771, "num_tokens": 363110876.0, "step": 9517 }, { "epoch": 1.2107874316244753, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.855663299560547, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8651275634765625, "num_tokens": 363155310.0, "step": 9518 }, { "epoch": 1.2109146419030659, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.881916046142578, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8786997199058533, "num_tokens": 363193166.0, "step": 9519 }, { "epoch": 1.2110418521816562, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.915552139282227, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.8500644564628601, "num_tokens": 363225163.0, "step": 9520 }, { "epoch": 1.2111690624602467, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.922728538513184, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8677346706390381, "num_tokens": 363264720.0, "step": 9521 }, { "epoch": 1.2112962727388372, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.828843116760254, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8531824350357056, "num_tokens": 363305029.0, "step": 9522 }, { "epoch": 1.2114234830174277, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.861077308654785, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8745813369750977, "num_tokens": 363344128.0, "step": 9523 }, { "epoch": 1.2115506932960183, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.906610488891602, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8670700192451477, "num_tokens": 363385903.0, "step": 9524 }, { "epoch": 1.2116779035746088, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.899975776672363, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8698967695236206, "num_tokens": 363425216.0, "step": 9525 }, { "epoch": 1.2118051138531993, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.89880084991455, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8789166808128357, "num_tokens": 363462691.0, "step": 9526 }, { "epoch": 1.2119323241317899, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.878179550170898, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8533832430839539, "num_tokens": 363495849.0, "step": 9527 }, { "epoch": 1.2120595344103804, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.949111938476562, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8676242232322693, "num_tokens": 363535268.0, "step": 9528 }, { "epoch": 1.212186744688971, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.921832084655762, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8501652479171753, "num_tokens": 363573523.0, "step": 9529 }, { "epoch": 1.2123139549675614, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.922018051147461, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8672444820404053, "num_tokens": 363607545.0, "step": 9530 }, { "epoch": 1.212441165246152, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.887762069702148, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8704076409339905, "num_tokens": 363648456.0, "step": 9531 }, { "epoch": 1.2125683755247425, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.902491569519043, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8768484592437744, "num_tokens": 363688726.0, "step": 9532 }, { "epoch": 1.212695585803333, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.94416332244873, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8608486652374268, "num_tokens": 363728557.0, "step": 9533 }, { "epoch": 1.2128227960819233, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.918438911437988, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8671246767044067, "num_tokens": 363764191.0, "step": 9534 }, { "epoch": 1.2129500063605139, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.982940673828125, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8566087484359741, "num_tokens": 363798126.0, "step": 9535 }, { "epoch": 1.2130772166391044, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.942093849182129, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8519758582115173, "num_tokens": 363836037.0, "step": 9536 }, { "epoch": 1.213204426917695, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.952079772949219, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8877504467964172, "num_tokens": 363872359.0, "step": 9537 }, { "epoch": 1.2133316371962855, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.879749298095703, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8616560697555542, "num_tokens": 363914206.0, "step": 9538 }, { "epoch": 1.213458847474876, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.963438034057617, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8689435124397278, "num_tokens": 363956147.0, "step": 9539 }, { "epoch": 1.2135860577534665, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.894432067871094, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8851481676101685, "num_tokens": 363986328.0, "step": 9540 }, { "epoch": 1.213713268032057, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.918083190917969, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8651612401008606, "num_tokens": 364024905.0, "step": 9541 }, { "epoch": 1.2138404783106476, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.906153678894043, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8686575889587402, "num_tokens": 364068628.0, "step": 9542 }, { "epoch": 1.213967688589238, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.923365592956543, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.864274263381958, "num_tokens": 364104261.0, "step": 9543 }, { "epoch": 1.2140948988678284, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.906094551086426, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8654324412345886, "num_tokens": 364147524.0, "step": 9544 }, { "epoch": 1.214222109146419, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.881195068359375, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8695148825645447, "num_tokens": 364187715.0, "step": 9545 }, { "epoch": 1.2143493194250095, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.963156700134277, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8784838914871216, "num_tokens": 364226081.0, "step": 9546 }, { "epoch": 1.2144765297036, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.851460456848145, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8686573505401611, "num_tokens": 364268390.0, "step": 9547 }, { "epoch": 1.2146037399821905, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.980913162231445, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.870142936706543, "num_tokens": 364310176.0, "step": 9548 }, { "epoch": 1.214730950260781, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.9071683883667, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8799367547035217, "num_tokens": 364345652.0, "step": 9549 }, { "epoch": 1.2148581605393716, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.911646842956543, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8638074398040771, "num_tokens": 364383810.0, "step": 9550 }, { "epoch": 1.214985370817962, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.898903846740723, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.871822714805603, "num_tokens": 364421169.0, "step": 9551 }, { "epoch": 1.2151125810965526, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.939562797546387, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8618700504302979, "num_tokens": 364452552.0, "step": 9552 }, { "epoch": 1.2152397913751432, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.911212921142578, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8659869432449341, "num_tokens": 364486779.0, "step": 9553 }, { "epoch": 1.2153670016537337, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.928380012512207, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8774507641792297, "num_tokens": 364526237.0, "step": 9554 }, { "epoch": 1.2154942119323242, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.989572525024414, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8746466636657715, "num_tokens": 364568446.0, "step": 9555 }, { "epoch": 1.2156214222109147, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.93504810333252, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8760285377502441, "num_tokens": 364604448.0, "step": 9556 }, { "epoch": 1.2157486324895053, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.989730834960938, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8653910160064697, "num_tokens": 364642707.0, "step": 9557 }, { "epoch": 1.2158758427680956, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.959555625915527, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.863491415977478, "num_tokens": 364678428.0, "step": 9558 }, { "epoch": 1.216003053046686, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.893959999084473, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8811241984367371, "num_tokens": 364718915.0, "step": 9559 }, { "epoch": 1.2161302633252766, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.91457462310791, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8844459056854248, "num_tokens": 364756030.0, "step": 9560 }, { "epoch": 1.2162574736038672, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.984404563903809, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8656508922576904, "num_tokens": 364796158.0, "step": 9561 }, { "epoch": 1.2163846838824577, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.999777793884277, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8759794235229492, "num_tokens": 364832722.0, "step": 9562 }, { "epoch": 1.2165118941610482, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.918917655944824, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8666011095046997, "num_tokens": 364874120.0, "step": 9563 }, { "epoch": 1.2166391044396387, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.983999252319336, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8698230981826782, "num_tokens": 364913166.0, "step": 9564 }, { "epoch": 1.2167663147182293, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.946733474731445, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8495191335678101, "num_tokens": 364949586.0, "step": 9565 }, { "epoch": 1.2168935249968198, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.96186351776123, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8508198261260986, "num_tokens": 364989070.0, "step": 9566 }, { "epoch": 1.2170207352754103, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.889532089233398, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8721373081207275, "num_tokens": 365027050.0, "step": 9567 }, { "epoch": 1.2171479455540009, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.93486499786377, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8680820465087891, "num_tokens": 365060050.0, "step": 9568 }, { "epoch": 1.2172751558325912, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.917671203613281, "learning_rate": 1e-06, "loss": 0.5023, "mean_token_accuracy": 0.8378016948699951, "num_tokens": 365098030.0, "step": 9569 }, { "epoch": 1.2174023661111817, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.892645835876465, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8700670003890991, "num_tokens": 365141143.0, "step": 9570 }, { "epoch": 1.2175295763897722, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.985528945922852, "learning_rate": 1e-06, "loss": 0.4978, "mean_token_accuracy": 0.8458420038223267, "num_tokens": 365174204.0, "step": 9571 }, { "epoch": 1.2176567866683627, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.94141960144043, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8687189817428589, "num_tokens": 365207170.0, "step": 9572 }, { "epoch": 1.2177839969469533, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.908249855041504, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.87132728099823, "num_tokens": 365248926.0, "step": 9573 }, { "epoch": 1.2179112072255438, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.99390697479248, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8714660406112671, "num_tokens": 365283588.0, "step": 9574 }, { "epoch": 1.2180384175041343, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.955106735229492, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8660635948181152, "num_tokens": 365324869.0, "step": 9575 }, { "epoch": 1.2181656277827249, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.926643371582031, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8638577461242676, "num_tokens": 365366832.0, "step": 9576 }, { "epoch": 1.2182928380613154, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.987258911132812, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8658955097198486, "num_tokens": 365401148.0, "step": 9577 }, { "epoch": 1.218420048339906, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.910338401794434, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.869462788105011, "num_tokens": 365442303.0, "step": 9578 }, { "epoch": 1.2185472586184964, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.91273307800293, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8802912831306458, "num_tokens": 365477877.0, "step": 9579 }, { "epoch": 1.218674468897087, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.896753311157227, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8564918041229248, "num_tokens": 365520648.0, "step": 9580 }, { "epoch": 1.2188016791756775, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.980280876159668, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8507992625236511, "num_tokens": 365555129.0, "step": 9581 }, { "epoch": 1.218928889454268, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.948408126831055, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8731306791305542, "num_tokens": 365594368.0, "step": 9582 }, { "epoch": 1.2190560997328583, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.855354309082031, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8677664995193481, "num_tokens": 365634763.0, "step": 9583 }, { "epoch": 1.2191833100114489, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.941873550415039, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8839842081069946, "num_tokens": 365675951.0, "step": 9584 }, { "epoch": 1.2193105202900394, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.95718002319336, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8755300045013428, "num_tokens": 365710348.0, "step": 9585 }, { "epoch": 1.21943773056863, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.913874626159668, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8706597089767456, "num_tokens": 365745166.0, "step": 9586 }, { "epoch": 1.2195649408472204, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.91907787322998, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8812509775161743, "num_tokens": 365780122.0, "step": 9587 }, { "epoch": 1.219692151125811, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.962425231933594, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8684786558151245, "num_tokens": 365816326.0, "step": 9588 }, { "epoch": 1.2198193614044015, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.879937171936035, "learning_rate": 1e-06, "loss": 0.4887, "mean_token_accuracy": 0.8419283628463745, "num_tokens": 365858938.0, "step": 9589 }, { "epoch": 1.219946571682992, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.998444557189941, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.857830286026001, "num_tokens": 365893892.0, "step": 9590 }, { "epoch": 1.2200737819615826, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.92910385131836, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.859019935131073, "num_tokens": 365938787.0, "step": 9591 }, { "epoch": 1.220200992240173, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.882328987121582, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8675292730331421, "num_tokens": 365972664.0, "step": 9592 }, { "epoch": 1.2203282025187634, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.018280982971191, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8741549253463745, "num_tokens": 366013367.0, "step": 9593 }, { "epoch": 1.220455412797354, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.897719383239746, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8740931153297424, "num_tokens": 366051637.0, "step": 9594 }, { "epoch": 1.2205826230759445, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.954856872558594, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8647225499153137, "num_tokens": 366092713.0, "step": 9595 }, { "epoch": 1.220709833354535, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.997026443481445, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.869012713432312, "num_tokens": 366131594.0, "step": 9596 }, { "epoch": 1.2208370436331255, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.895730018615723, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.863002359867096, "num_tokens": 366170871.0, "step": 9597 }, { "epoch": 1.220964253911716, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.934560775756836, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8572349548339844, "num_tokens": 366212974.0, "step": 9598 }, { "epoch": 1.2210914641903066, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.94609260559082, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8596460819244385, "num_tokens": 366255904.0, "step": 9599 }, { "epoch": 1.221218674468897, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.969258308410645, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8732013702392578, "num_tokens": 366288887.0, "step": 9600 }, { "epoch": 1.2213458847474876, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.971579551696777, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8810793161392212, "num_tokens": 366328796.0, "step": 9601 }, { "epoch": 1.2214730950260781, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.978337287902832, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8640486001968384, "num_tokens": 366373093.0, "step": 9602 }, { "epoch": 1.2216003053046687, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.975296974182129, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8768330812454224, "num_tokens": 366414353.0, "step": 9603 }, { "epoch": 1.2217275155832592, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.908445358276367, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8806067705154419, "num_tokens": 366450683.0, "step": 9604 }, { "epoch": 1.2218547258618497, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.98318099975586, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8730826377868652, "num_tokens": 366487158.0, "step": 9605 }, { "epoch": 1.2219819361404403, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.907858848571777, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8698079586029053, "num_tokens": 366529984.0, "step": 9606 }, { "epoch": 1.2221091464190306, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.944587707519531, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8714330792427063, "num_tokens": 366564100.0, "step": 9607 }, { "epoch": 1.222236356697621, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.991276741027832, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8753504753112793, "num_tokens": 366603752.0, "step": 9608 }, { "epoch": 1.2223635669762116, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.866169929504395, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8696001172065735, "num_tokens": 366642170.0, "step": 9609 }, { "epoch": 1.2224907772548022, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.98982048034668, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8547354936599731, "num_tokens": 366684038.0, "step": 9610 }, { "epoch": 1.2226179875333927, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.974881172180176, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8656805753707886, "num_tokens": 366717832.0, "step": 9611 }, { "epoch": 1.2227451978119832, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.94505786895752, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8623256683349609, "num_tokens": 366757999.0, "step": 9612 }, { "epoch": 1.2228724080905737, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.882113456726074, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8771004676818848, "num_tokens": 366798533.0, "step": 9613 }, { "epoch": 1.2229996183691643, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.000334739685059, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8654308319091797, "num_tokens": 366843066.0, "step": 9614 }, { "epoch": 1.2231268286477548, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.895926475524902, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8835257291793823, "num_tokens": 366883190.0, "step": 9615 }, { "epoch": 1.2232540389263453, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.000061988830566, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8719999194145203, "num_tokens": 366920642.0, "step": 9616 }, { "epoch": 1.2233812492049359, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.031538009643555, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8692057728767395, "num_tokens": 366954468.0, "step": 9617 }, { "epoch": 1.2235084594835262, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.911645889282227, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.852091908454895, "num_tokens": 366991170.0, "step": 9618 }, { "epoch": 1.2236356697621167, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.960792541503906, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8731946349143982, "num_tokens": 367031833.0, "step": 9619 }, { "epoch": 1.2237628800407072, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.948198318481445, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8635193109512329, "num_tokens": 367074709.0, "step": 9620 }, { "epoch": 1.2238900903192977, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.990225791931152, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8645495176315308, "num_tokens": 367106820.0, "step": 9621 }, { "epoch": 1.2240173005978883, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.860568046569824, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8819260001182556, "num_tokens": 367142298.0, "step": 9622 }, { "epoch": 1.2241445108764788, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.006465911865234, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8586363792419434, "num_tokens": 367179957.0, "step": 9623 }, { "epoch": 1.2242717211550693, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.86133098602295, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.870057225227356, "num_tokens": 367226800.0, "step": 9624 }, { "epoch": 1.2243989314336599, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.945603370666504, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8716444969177246, "num_tokens": 367257168.0, "step": 9625 }, { "epoch": 1.2245261417122504, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.975095748901367, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8705537915229797, "num_tokens": 367295254.0, "step": 9626 }, { "epoch": 1.224653351990841, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.939428329467773, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8557060360908508, "num_tokens": 367335814.0, "step": 9627 }, { "epoch": 1.2247805622694314, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.927969932556152, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8646219372749329, "num_tokens": 367377592.0, "step": 9628 }, { "epoch": 1.224907772548022, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.967229843139648, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8607287406921387, "num_tokens": 367408069.0, "step": 9629 }, { "epoch": 1.2250349828266125, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.957509994506836, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8693147897720337, "num_tokens": 367444787.0, "step": 9630 }, { "epoch": 1.225162193105203, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.84582805633545, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8542888760566711, "num_tokens": 367486848.0, "step": 9631 }, { "epoch": 1.2252894033837933, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.938372611999512, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.878299355506897, "num_tokens": 367530456.0, "step": 9632 }, { "epoch": 1.2254166136623839, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.945013999938965, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8708663582801819, "num_tokens": 367566513.0, "step": 9633 }, { "epoch": 1.2255438239409744, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.978363990783691, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8481552004814148, "num_tokens": 367607757.0, "step": 9634 }, { "epoch": 1.225671034219565, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.975515365600586, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8701981902122498, "num_tokens": 367645525.0, "step": 9635 }, { "epoch": 1.2257982444981554, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.05217456817627, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8766484260559082, "num_tokens": 367680840.0, "step": 9636 }, { "epoch": 1.225925454776746, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.937944412231445, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.875871479511261, "num_tokens": 367717444.0, "step": 9637 }, { "epoch": 1.2260526650553365, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.97976016998291, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8464823961257935, "num_tokens": 367756548.0, "step": 9638 }, { "epoch": 1.226179875333927, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.895660400390625, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8623769283294678, "num_tokens": 367798341.0, "step": 9639 }, { "epoch": 1.2263070856125176, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.913712501525879, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8749887943267822, "num_tokens": 367835087.0, "step": 9640 }, { "epoch": 1.226434295891108, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.900824546813965, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8697527647018433, "num_tokens": 367873816.0, "step": 9641 }, { "epoch": 1.2265615061696984, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.961052894592285, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8639042377471924, "num_tokens": 367913621.0, "step": 9642 }, { "epoch": 1.226688716448289, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.937187194824219, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8588420152664185, "num_tokens": 367946881.0, "step": 9643 }, { "epoch": 1.2268159267268794, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.985942840576172, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8718764185905457, "num_tokens": 367980806.0, "step": 9644 }, { "epoch": 1.22694313700547, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.008816719055176, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8745914697647095, "num_tokens": 368016041.0, "step": 9645 }, { "epoch": 1.2270703472840605, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.928731918334961, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8653485774993896, "num_tokens": 368052331.0, "step": 9646 }, { "epoch": 1.227197557562651, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.016745567321777, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8770188093185425, "num_tokens": 368091458.0, "step": 9647 }, { "epoch": 1.2273247678412416, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.886253356933594, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8562273979187012, "num_tokens": 368132433.0, "step": 9648 }, { "epoch": 1.227451978119832, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.027846336364746, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8685672283172607, "num_tokens": 368171030.0, "step": 9649 }, { "epoch": 1.2275791883984226, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.910017967224121, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8568847179412842, "num_tokens": 368211555.0, "step": 9650 }, { "epoch": 1.2277063986770131, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.94054889678955, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.874381422996521, "num_tokens": 368252076.0, "step": 9651 }, { "epoch": 1.2278336089556037, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.942748069763184, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8616050481796265, "num_tokens": 368291265.0, "step": 9652 }, { "epoch": 1.2279608192341942, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.993284225463867, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8602427244186401, "num_tokens": 368332178.0, "step": 9653 }, { "epoch": 1.2280880295127847, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.9756498336792, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8655398488044739, "num_tokens": 368374103.0, "step": 9654 }, { "epoch": 1.2282152397913753, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.945440292358398, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.870031476020813, "num_tokens": 368411928.0, "step": 9655 }, { "epoch": 1.2283424500699656, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.974753379821777, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8444882035255432, "num_tokens": 368449200.0, "step": 9656 }, { "epoch": 1.228469660348556, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.88042163848877, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8503994941711426, "num_tokens": 368489935.0, "step": 9657 }, { "epoch": 1.2285968706271466, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.920513153076172, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8812638521194458, "num_tokens": 368522227.0, "step": 9658 }, { "epoch": 1.2287240809057371, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.952261924743652, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8676578402519226, "num_tokens": 368555084.0, "step": 9659 }, { "epoch": 1.2288512911843277, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.908632278442383, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8634570837020874, "num_tokens": 368591845.0, "step": 9660 }, { "epoch": 1.2289785014629182, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.913676261901855, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8574222326278687, "num_tokens": 368637841.0, "step": 9661 }, { "epoch": 1.2291057117415087, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.931809425354004, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8822610378265381, "num_tokens": 368677470.0, "step": 9662 }, { "epoch": 1.2292329220200993, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.955406188964844, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8811757564544678, "num_tokens": 368718412.0, "step": 9663 }, { "epoch": 1.2293601322986898, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.917987823486328, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8563956022262573, "num_tokens": 368759267.0, "step": 9664 }, { "epoch": 1.2294873425772803, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.995835304260254, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8594653606414795, "num_tokens": 368797586.0, "step": 9665 }, { "epoch": 1.2296145528558708, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.98371696472168, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8597525358200073, "num_tokens": 368833801.0, "step": 9666 }, { "epoch": 1.2297417631344612, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.975849151611328, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8518460988998413, "num_tokens": 368878513.0, "step": 9667 }, { "epoch": 1.2298689734130517, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.963112831115723, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8604843616485596, "num_tokens": 368917815.0, "step": 9668 }, { "epoch": 1.2299961836916422, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.969488143920898, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8567472696304321, "num_tokens": 368954971.0, "step": 9669 }, { "epoch": 1.2301233939702327, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.97546100616455, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8722857236862183, "num_tokens": 368996212.0, "step": 9670 }, { "epoch": 1.2302506042488233, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.951433181762695, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8809331059455872, "num_tokens": 369032186.0, "step": 9671 }, { "epoch": 1.2303778145274138, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.959235191345215, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8720371723175049, "num_tokens": 369077565.0, "step": 9672 }, { "epoch": 1.2305050248060043, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.023636817932129, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8737574219703674, "num_tokens": 369111935.0, "step": 9673 }, { "epoch": 1.2306322350845948, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.960061073303223, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.86484694480896, "num_tokens": 369154374.0, "step": 9674 }, { "epoch": 1.2307594453631854, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.947125434875488, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8587714433670044, "num_tokens": 369198557.0, "step": 9675 }, { "epoch": 1.230886655641776, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.041034698486328, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8507946729660034, "num_tokens": 369237756.0, "step": 9676 }, { "epoch": 1.2310138659203664, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.987687110900879, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8542970418930054, "num_tokens": 369270155.0, "step": 9677 }, { "epoch": 1.231141076198957, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.964163780212402, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8591850996017456, "num_tokens": 369310870.0, "step": 9678 }, { "epoch": 1.2312682864775475, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.010058403015137, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8490716218948364, "num_tokens": 369344679.0, "step": 9679 }, { "epoch": 1.231395496756138, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.010540008544922, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8672329187393188, "num_tokens": 369380388.0, "step": 9680 }, { "epoch": 1.2315227070347283, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.944673538208008, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8788648843765259, "num_tokens": 369422327.0, "step": 9681 }, { "epoch": 1.2316499173133189, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.997791290283203, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8731815814971924, "num_tokens": 369457160.0, "step": 9682 }, { "epoch": 1.2317771275919094, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.964972496032715, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8670809268951416, "num_tokens": 369497883.0, "step": 9683 }, { "epoch": 1.2319043378705, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.963493347167969, "learning_rate": 1e-06, "loss": 0.3234, "mean_token_accuracy": 0.896218478679657, "num_tokens": 369540014.0, "step": 9684 }, { "epoch": 1.2320315481490904, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.907366752624512, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8661607503890991, "num_tokens": 369580424.0, "step": 9685 }, { "epoch": 1.232158758427681, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.965771675109863, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8637166023254395, "num_tokens": 369623615.0, "step": 9686 }, { "epoch": 1.2322859687062715, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.991498947143555, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8552241325378418, "num_tokens": 369658408.0, "step": 9687 }, { "epoch": 1.232413178984862, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.91060733795166, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8644160032272339, "num_tokens": 369696311.0, "step": 9688 }, { "epoch": 1.2325403892634526, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.972465515136719, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8543172478675842, "num_tokens": 369737473.0, "step": 9689 }, { "epoch": 1.232667599542043, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.913240432739258, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8646274209022522, "num_tokens": 369770151.0, "step": 9690 }, { "epoch": 1.2327948098206334, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.950119972229004, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8856002688407898, "num_tokens": 369813503.0, "step": 9691 }, { "epoch": 1.232922020099224, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.978333473205566, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8688579201698303, "num_tokens": 369848768.0, "step": 9692 }, { "epoch": 1.2330492303778144, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.983187675476074, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8698858022689819, "num_tokens": 369892510.0, "step": 9693 }, { "epoch": 1.233176440656405, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.956754684448242, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8746188879013062, "num_tokens": 369929638.0, "step": 9694 }, { "epoch": 1.2333036509349955, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.90250301361084, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8876197338104248, "num_tokens": 369971201.0, "step": 9695 }, { "epoch": 1.233430861213586, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.027999877929688, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8615071177482605, "num_tokens": 370011713.0, "step": 9696 }, { "epoch": 1.2335580714921766, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.974991798400879, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8470808267593384, "num_tokens": 370042647.0, "step": 9697 }, { "epoch": 1.233685281770767, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.012067794799805, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8652390241622925, "num_tokens": 370083976.0, "step": 9698 }, { "epoch": 1.2338124920493576, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.97638988494873, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8636080622673035, "num_tokens": 370126435.0, "step": 9699 }, { "epoch": 1.2339397023279481, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.945737838745117, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.876907467842102, "num_tokens": 370165038.0, "step": 9700 }, { "epoch": 1.2340669126065387, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.049347877502441, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8592731952667236, "num_tokens": 370203438.0, "step": 9701 }, { "epoch": 1.2341941228851292, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.969133377075195, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8816897869110107, "num_tokens": 370235144.0, "step": 9702 }, { "epoch": 1.2343213331637197, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.969258308410645, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8617315292358398, "num_tokens": 370278936.0, "step": 9703 }, { "epoch": 1.2344485434423103, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.943849563598633, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8710049390792847, "num_tokens": 370312595.0, "step": 9704 }, { "epoch": 1.2345757537209006, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.981656074523926, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8642566204071045, "num_tokens": 370351697.0, "step": 9705 }, { "epoch": 1.234702963999491, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.939045906066895, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8803387880325317, "num_tokens": 370391264.0, "step": 9706 }, { "epoch": 1.2348301742780816, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.022120475769043, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8744372129440308, "num_tokens": 370425248.0, "step": 9707 }, { "epoch": 1.2349573845566721, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.98369026184082, "learning_rate": 1e-06, "loss": 0.4756, "mean_token_accuracy": 0.8457971811294556, "num_tokens": 370462670.0, "step": 9708 }, { "epoch": 1.2350845948352627, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.008831024169922, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8479729890823364, "num_tokens": 370506126.0, "step": 9709 }, { "epoch": 1.2352118051138532, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.976595878601074, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8779122233390808, "num_tokens": 370546895.0, "step": 9710 }, { "epoch": 1.2353390153924437, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.054207801818848, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8712179660797119, "num_tokens": 370590645.0, "step": 9711 }, { "epoch": 1.2354662256710343, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.999496459960938, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8703110814094543, "num_tokens": 370628614.0, "step": 9712 }, { "epoch": 1.2355934359496248, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.018877029418945, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8690011501312256, "num_tokens": 370666082.0, "step": 9713 }, { "epoch": 1.2357206462282153, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.021162033081055, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8529729247093201, "num_tokens": 370705780.0, "step": 9714 }, { "epoch": 1.2358478565068058, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.958903312683105, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8730229139328003, "num_tokens": 370742550.0, "step": 9715 }, { "epoch": 1.2359750667853961, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.032501220703125, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8717308044433594, "num_tokens": 370784406.0, "step": 9716 }, { "epoch": 1.2361022770639867, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.927824020385742, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8621605634689331, "num_tokens": 370820911.0, "step": 9717 }, { "epoch": 1.2362294873425772, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.050592422485352, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8783722519874573, "num_tokens": 370857427.0, "step": 9718 }, { "epoch": 1.2363566976211677, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.955865859985352, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8571305274963379, "num_tokens": 370897989.0, "step": 9719 }, { "epoch": 1.2364839078997583, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.009398460388184, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8543060421943665, "num_tokens": 370932672.0, "step": 9720 }, { "epoch": 1.2366111181783488, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.049779891967773, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.858614444732666, "num_tokens": 370971027.0, "step": 9721 }, { "epoch": 1.2367383284569393, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.902229309082031, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8784116506576538, "num_tokens": 371014742.0, "step": 9722 }, { "epoch": 1.2368655387355298, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.96668529510498, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8506325483322144, "num_tokens": 371052065.0, "step": 9723 }, { "epoch": 1.2369927490141204, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.967345237731934, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8457798361778259, "num_tokens": 371093897.0, "step": 9724 }, { "epoch": 1.237119959292711, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.92257308959961, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8672347068786621, "num_tokens": 371132390.0, "step": 9725 }, { "epoch": 1.2372471695713014, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.911903381347656, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8783859014511108, "num_tokens": 371170468.0, "step": 9726 }, { "epoch": 1.237374379849892, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.968585968017578, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8646256327629089, "num_tokens": 371209323.0, "step": 9727 }, { "epoch": 1.2375015901284825, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.927594184875488, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8663422465324402, "num_tokens": 371246526.0, "step": 9728 }, { "epoch": 1.237628800407073, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.95517349243164, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8710671067237854, "num_tokens": 371282850.0, "step": 9729 }, { "epoch": 1.2377560106856633, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.977739334106445, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8757772445678711, "num_tokens": 371312614.0, "step": 9730 }, { "epoch": 1.2378832209642538, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.996541023254395, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8685747385025024, "num_tokens": 371350170.0, "step": 9731 }, { "epoch": 1.2380104312428444, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.92780876159668, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8744825720787048, "num_tokens": 371389295.0, "step": 9732 }, { "epoch": 1.238137641521435, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.962776184082031, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8670936822891235, "num_tokens": 371430579.0, "step": 9733 }, { "epoch": 1.2382648518000254, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.953824996948242, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8581105470657349, "num_tokens": 371465087.0, "step": 9734 }, { "epoch": 1.238392062078616, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.92286205291748, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8745677471160889, "num_tokens": 371503549.0, "step": 9735 }, { "epoch": 1.2385192723572065, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.994975090026855, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8866637349128723, "num_tokens": 371542953.0, "step": 9736 }, { "epoch": 1.238646482635797, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.90267562866211, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8716358542442322, "num_tokens": 371582541.0, "step": 9737 }, { "epoch": 1.2387736929143875, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.94465446472168, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.877510666847229, "num_tokens": 371617221.0, "step": 9738 }, { "epoch": 1.238900903192978, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.933712005615234, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8703303337097168, "num_tokens": 371660773.0, "step": 9739 }, { "epoch": 1.2390281134715684, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.98699951171875, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8621468544006348, "num_tokens": 371694763.0, "step": 9740 }, { "epoch": 1.239155323750159, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.985825538635254, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8669811487197876, "num_tokens": 371734793.0, "step": 9741 }, { "epoch": 1.2392825340287494, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.984590530395508, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8619198799133301, "num_tokens": 371775691.0, "step": 9742 }, { "epoch": 1.23940974430734, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.023682594299316, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8600454926490784, "num_tokens": 371817319.0, "step": 9743 }, { "epoch": 1.2395369545859305, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.953926086425781, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8670911192893982, "num_tokens": 371847431.0, "step": 9744 }, { "epoch": 1.239664164864521, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.9467191696167, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8802130818367004, "num_tokens": 371881788.0, "step": 9745 }, { "epoch": 1.2397913751431116, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.000332832336426, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8631691336631775, "num_tokens": 371917388.0, "step": 9746 }, { "epoch": 1.239918585421702, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.972238540649414, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8759492635726929, "num_tokens": 371957561.0, "step": 9747 }, { "epoch": 1.2400457957002926, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.037641525268555, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8611881136894226, "num_tokens": 371987322.0, "step": 9748 }, { "epoch": 1.2401730059788831, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.016922950744629, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8698815107345581, "num_tokens": 372027263.0, "step": 9749 }, { "epoch": 1.2403002162574737, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.003654479980469, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8719176054000854, "num_tokens": 372061441.0, "step": 9750 }, { "epoch": 1.2404274265360642, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.976873397827148, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8729899525642395, "num_tokens": 372099085.0, "step": 9751 }, { "epoch": 1.2405546368146547, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.937066078186035, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8703882694244385, "num_tokens": 372134106.0, "step": 9752 }, { "epoch": 1.2406818470932452, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.070605278015137, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8764696717262268, "num_tokens": 372164635.0, "step": 9753 }, { "epoch": 1.2408090573718356, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.900938987731934, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8739009499549866, "num_tokens": 372204924.0, "step": 9754 }, { "epoch": 1.240936267650426, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.969966888427734, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8673542737960815, "num_tokens": 372240939.0, "step": 9755 }, { "epoch": 1.2410634779290166, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.98583698272705, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8634119033813477, "num_tokens": 372272355.0, "step": 9756 }, { "epoch": 1.2411906882076071, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.012419700622559, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8659823536872864, "num_tokens": 372309566.0, "step": 9757 }, { "epoch": 1.2413178984861977, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.972107887268066, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8624981045722961, "num_tokens": 372347768.0, "step": 9758 }, { "epoch": 1.2414451087647882, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.969877243041992, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8760980367660522, "num_tokens": 372381020.0, "step": 9759 }, { "epoch": 1.2415723190433787, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.965184211730957, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.856545627117157, "num_tokens": 372422434.0, "step": 9760 }, { "epoch": 1.2416995293219693, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.982898712158203, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8594853281974792, "num_tokens": 372459342.0, "step": 9761 }, { "epoch": 1.2418267396005598, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.946783065795898, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8708972930908203, "num_tokens": 372496637.0, "step": 9762 }, { "epoch": 1.2419539498791503, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.949481964111328, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8799594044685364, "num_tokens": 372532068.0, "step": 9763 }, { "epoch": 1.2420811601577408, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.06186294555664, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.874252438545227, "num_tokens": 372565460.0, "step": 9764 }, { "epoch": 1.2422083704363311, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.994117736816406, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.871268630027771, "num_tokens": 372605998.0, "step": 9765 }, { "epoch": 1.2423355807149217, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.919621467590332, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8572399020195007, "num_tokens": 372642038.0, "step": 9766 }, { "epoch": 1.2424627909935122, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.985671043395996, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8643777966499329, "num_tokens": 372678192.0, "step": 9767 }, { "epoch": 1.2425900012721027, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.930500984191895, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8638460636138916, "num_tokens": 372712867.0, "step": 9768 }, { "epoch": 1.2427172115506933, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.043401718139648, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8502891063690186, "num_tokens": 372745790.0, "step": 9769 }, { "epoch": 1.2428444218292838, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.029565811157227, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8789215087890625, "num_tokens": 372777634.0, "step": 9770 }, { "epoch": 1.2429716321078743, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.980902671813965, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8525753021240234, "num_tokens": 372820527.0, "step": 9771 }, { "epoch": 1.2430988423864648, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.989697456359863, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8775080442428589, "num_tokens": 372858479.0, "step": 9772 }, { "epoch": 1.2432260526650554, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.96809196472168, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8745430707931519, "num_tokens": 372895219.0, "step": 9773 }, { "epoch": 1.243353262943646, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.95029354095459, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8799905776977539, "num_tokens": 372934491.0, "step": 9774 }, { "epoch": 1.2434804732222364, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.962625503540039, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8592068552970886, "num_tokens": 372972892.0, "step": 9775 }, { "epoch": 1.243607683500827, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.035001754760742, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8708095550537109, "num_tokens": 373011564.0, "step": 9776 }, { "epoch": 1.2437348937794175, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.01396369934082, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8694584369659424, "num_tokens": 373049444.0, "step": 9777 }, { "epoch": 1.243862104058008, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.958213806152344, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8666588068008423, "num_tokens": 373085466.0, "step": 9778 }, { "epoch": 1.2439893143365983, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.979597091674805, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8456002473831177, "num_tokens": 373124712.0, "step": 9779 }, { "epoch": 1.2441165246151888, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.975213050842285, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8731472492218018, "num_tokens": 373160008.0, "step": 9780 }, { "epoch": 1.2442437348937794, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.040191650390625, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8688903450965881, "num_tokens": 373192110.0, "step": 9781 }, { "epoch": 1.24437094517237, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.985458374023438, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8692721128463745, "num_tokens": 373234432.0, "step": 9782 }, { "epoch": 1.2444981554509604, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.997798919677734, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8582346439361572, "num_tokens": 373270725.0, "step": 9783 }, { "epoch": 1.244625365729551, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.89964771270752, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.846243679523468, "num_tokens": 373307531.0, "step": 9784 }, { "epoch": 1.2447525760081415, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.032026290893555, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8717705607414246, "num_tokens": 373351657.0, "step": 9785 }, { "epoch": 1.244879786286732, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.962860107421875, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8736940622329712, "num_tokens": 373388811.0, "step": 9786 }, { "epoch": 1.2450069965653225, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.962348937988281, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8769676089286804, "num_tokens": 373423378.0, "step": 9787 }, { "epoch": 1.245134206843913, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.994997024536133, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8696432113647461, "num_tokens": 373460982.0, "step": 9788 }, { "epoch": 1.2452614171225034, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.924077033996582, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8743094801902771, "num_tokens": 373501474.0, "step": 9789 }, { "epoch": 1.245388627401094, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.969733238220215, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8714351058006287, "num_tokens": 373537830.0, "step": 9790 }, { "epoch": 1.2455158376796844, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.969986915588379, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8609137535095215, "num_tokens": 373573975.0, "step": 9791 }, { "epoch": 1.245643047958275, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.982110977172852, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8479152321815491, "num_tokens": 373612157.0, "step": 9792 }, { "epoch": 1.2457702582368655, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.988697052001953, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8874481916427612, "num_tokens": 373645065.0, "step": 9793 }, { "epoch": 1.245897468515456, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.948054313659668, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8626559972763062, "num_tokens": 373687658.0, "step": 9794 }, { "epoch": 1.2460246787940465, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.990169525146484, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8692556619644165, "num_tokens": 373732262.0, "step": 9795 }, { "epoch": 1.246151889072637, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.964005470275879, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8484534025192261, "num_tokens": 373772410.0, "step": 9796 }, { "epoch": 1.2462790993512276, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.053624153137207, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8683905005455017, "num_tokens": 373808773.0, "step": 9797 }, { "epoch": 1.2464063096298181, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.950417518615723, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8630648851394653, "num_tokens": 373848945.0, "step": 9798 }, { "epoch": 1.2465335199084087, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.011406898498535, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8630580902099609, "num_tokens": 373892716.0, "step": 9799 }, { "epoch": 1.2466607301869992, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.966489791870117, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8832485675811768, "num_tokens": 373929443.0, "step": 9800 }, { "epoch": 1.2467879404655897, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.00554370880127, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.868473470211029, "num_tokens": 373968882.0, "step": 9801 }, { "epoch": 1.2469151507441802, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.951277732849121, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8805256485939026, "num_tokens": 374009265.0, "step": 9802 }, { "epoch": 1.2470423610227706, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.008641242980957, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8529639840126038, "num_tokens": 374047152.0, "step": 9803 }, { "epoch": 1.247169571301361, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.003284454345703, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8647512197494507, "num_tokens": 374083127.0, "step": 9804 }, { "epoch": 1.2472967815799516, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.948530197143555, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.876744270324707, "num_tokens": 374127086.0, "step": 9805 }, { "epoch": 1.2474239918585421, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.009571075439453, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8617255687713623, "num_tokens": 374163879.0, "step": 9806 }, { "epoch": 1.2475512021371327, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.98691177368164, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8717189431190491, "num_tokens": 374206639.0, "step": 9807 }, { "epoch": 1.2476784124157232, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.036331176757812, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8648377656936646, "num_tokens": 374244106.0, "step": 9808 }, { "epoch": 1.2478056226943137, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.072311401367188, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8644556999206543, "num_tokens": 374285776.0, "step": 9809 }, { "epoch": 1.2479328329729042, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.981413841247559, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8776168823242188, "num_tokens": 374323036.0, "step": 9810 }, { "epoch": 1.2480600432514948, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.055596351623535, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8746334910392761, "num_tokens": 374360204.0, "step": 9811 }, { "epoch": 1.2481872535300853, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.976698875427246, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8775614500045776, "num_tokens": 374394961.0, "step": 9812 }, { "epoch": 1.2483144638086758, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.040209770202637, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8694815039634705, "num_tokens": 374436142.0, "step": 9813 }, { "epoch": 1.2484416740872661, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.044161796569824, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8646476864814758, "num_tokens": 374474263.0, "step": 9814 }, { "epoch": 1.2485688843658567, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.998941421508789, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8626026511192322, "num_tokens": 374510250.0, "step": 9815 }, { "epoch": 1.2486960946444472, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.010265350341797, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8854576349258423, "num_tokens": 374549579.0, "step": 9816 }, { "epoch": 1.2488233049230377, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.011557579040527, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8660376667976379, "num_tokens": 374588631.0, "step": 9817 }, { "epoch": 1.2489505152016283, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.929963111877441, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8635101318359375, "num_tokens": 374627092.0, "step": 9818 }, { "epoch": 1.2490777254802188, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.067251205444336, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.873184084892273, "num_tokens": 374657797.0, "step": 9819 }, { "epoch": 1.2492049357588093, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.068230628967285, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8728276491165161, "num_tokens": 374694011.0, "step": 9820 }, { "epoch": 1.2493321460373998, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.984049797058105, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8697604537010193, "num_tokens": 374735127.0, "step": 9821 }, { "epoch": 1.2494593563159904, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.983137130737305, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8657345771789551, "num_tokens": 374774170.0, "step": 9822 }, { "epoch": 1.249586566594581, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.083888053894043, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8689868450164795, "num_tokens": 374809162.0, "step": 9823 }, { "epoch": 1.2497137768731714, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.015907287597656, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8589472770690918, "num_tokens": 374847974.0, "step": 9824 }, { "epoch": 1.249840987151762, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.001173973083496, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8844261765480042, "num_tokens": 374894911.0, "step": 9825 }, { "epoch": 1.2499681974303525, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.016647338867188, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8633943796157837, "num_tokens": 374933534.0, "step": 9826 }, { "epoch": 1.250095407708943, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.082475662231445, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.855995237827301, "num_tokens": 374968183.0, "step": 9827 }, { "epoch": 1.2502226179875333, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.997241020202637, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8609033823013306, "num_tokens": 375005973.0, "step": 9828 }, { "epoch": 1.2503498282661238, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.035436630249023, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8609344959259033, "num_tokens": 375043446.0, "step": 9829 }, { "epoch": 1.2504770385447144, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.063700675964355, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8704713582992554, "num_tokens": 375081518.0, "step": 9830 }, { "epoch": 1.250604248823305, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.971263885498047, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8609280586242676, "num_tokens": 375117767.0, "step": 9831 }, { "epoch": 1.2507314591018954, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.035073280334473, "learning_rate": 1e-06, "loss": 0.4847, "mean_token_accuracy": 0.8440263867378235, "num_tokens": 375152061.0, "step": 9832 }, { "epoch": 1.250858669380486, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.968877792358398, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.861536979675293, "num_tokens": 375195677.0, "step": 9833 }, { "epoch": 1.2509858796590765, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.989294052124023, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8608567118644714, "num_tokens": 375230319.0, "step": 9834 }, { "epoch": 1.251113089937667, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.050177574157715, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8718697428703308, "num_tokens": 375268970.0, "step": 9835 }, { "epoch": 1.2512403002162575, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.041465759277344, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8739494681358337, "num_tokens": 375303661.0, "step": 9836 }, { "epoch": 1.2513675104948478, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.009730339050293, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8743815422058105, "num_tokens": 375343278.0, "step": 9837 }, { "epoch": 1.2514947207734384, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.973955154418945, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8733079433441162, "num_tokens": 375378574.0, "step": 9838 }, { "epoch": 1.251621931052029, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.07070541381836, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8623872399330139, "num_tokens": 375412781.0, "step": 9839 }, { "epoch": 1.2517491413306194, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.958744049072266, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8801568746566772, "num_tokens": 375450431.0, "step": 9840 }, { "epoch": 1.25187635160921, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.012068748474121, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8613453507423401, "num_tokens": 375490692.0, "step": 9841 }, { "epoch": 1.2520035618878005, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.977901458740234, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8727367520332336, "num_tokens": 375535278.0, "step": 9842 }, { "epoch": 1.252130772166391, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.03459358215332, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8613730669021606, "num_tokens": 375575859.0, "step": 9843 }, { "epoch": 1.2522579824449815, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.988961219787598, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8602844476699829, "num_tokens": 375611084.0, "step": 9844 }, { "epoch": 1.252385192723572, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.011805534362793, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8777467012405396, "num_tokens": 375644926.0, "step": 9845 }, { "epoch": 1.2525124030021626, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.98708724975586, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8767264485359192, "num_tokens": 375686970.0, "step": 9846 }, { "epoch": 1.2526396132807531, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.94003963470459, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8786982297897339, "num_tokens": 375727372.0, "step": 9847 }, { "epoch": 1.2527668235593437, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.969627380371094, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8566608428955078, "num_tokens": 375763568.0, "step": 9848 }, { "epoch": 1.2528940338379342, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.054736137390137, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8511099219322205, "num_tokens": 375797954.0, "step": 9849 }, { "epoch": 1.2530212441165247, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.974777221679688, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8659957051277161, "num_tokens": 375836362.0, "step": 9850 }, { "epoch": 1.2531484543951152, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.987168312072754, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8624709844589233, "num_tokens": 375879403.0, "step": 9851 }, { "epoch": 1.2532756646737058, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.045999526977539, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.870415210723877, "num_tokens": 375916728.0, "step": 9852 }, { "epoch": 1.253402874952296, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.041169166564941, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8613568544387817, "num_tokens": 375951743.0, "step": 9853 }, { "epoch": 1.2535300852308866, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.05881404876709, "learning_rate": 1e-06, "loss": 0.4893, "mean_token_accuracy": 0.8427133560180664, "num_tokens": 375988872.0, "step": 9854 }, { "epoch": 1.2536572955094771, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.966177940368652, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8570639491081238, "num_tokens": 376030388.0, "step": 9855 }, { "epoch": 1.2537845057880677, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.025748252868652, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8763300180435181, "num_tokens": 376062975.0, "step": 9856 }, { "epoch": 1.2539117160666582, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.965866088867188, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8600270748138428, "num_tokens": 376101873.0, "step": 9857 }, { "epoch": 1.2540389263452487, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.045075416564941, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8466111421585083, "num_tokens": 376138774.0, "step": 9858 }, { "epoch": 1.2541661366238392, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.976598739624023, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8591588735580444, "num_tokens": 376172357.0, "step": 9859 }, { "epoch": 1.2542933469024298, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.99488639831543, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8566871285438538, "num_tokens": 376212977.0, "step": 9860 }, { "epoch": 1.2544205571810203, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.990596771240234, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8569521903991699, "num_tokens": 376256989.0, "step": 9861 }, { "epoch": 1.2545477674596106, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.054182052612305, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.871993899345398, "num_tokens": 376300130.0, "step": 9862 }, { "epoch": 1.2546749777382011, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.033987998962402, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8703148365020752, "num_tokens": 376337475.0, "step": 9863 }, { "epoch": 1.2548021880167917, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.006552696228027, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.858684778213501, "num_tokens": 376376485.0, "step": 9864 }, { "epoch": 1.2549293982953822, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.915385246276855, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8734339475631714, "num_tokens": 376413393.0, "step": 9865 }, { "epoch": 1.2550566085739727, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.050143241882324, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8582838177680969, "num_tokens": 376448192.0, "step": 9866 }, { "epoch": 1.2551838188525632, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.99463939666748, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8704442977905273, "num_tokens": 376489382.0, "step": 9867 }, { "epoch": 1.2553110291311538, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.052704811096191, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8712304830551147, "num_tokens": 376527221.0, "step": 9868 }, { "epoch": 1.2554382394097443, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.943056106567383, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8623945713043213, "num_tokens": 376567045.0, "step": 9869 }, { "epoch": 1.2555654496883348, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.031089782714844, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8742554783821106, "num_tokens": 376607039.0, "step": 9870 }, { "epoch": 1.2556926599669254, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.093904495239258, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8709225058555603, "num_tokens": 376643220.0, "step": 9871 }, { "epoch": 1.255819870245516, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.974224090576172, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.886203944683075, "num_tokens": 376681417.0, "step": 9872 }, { "epoch": 1.2559470805241064, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.992029190063477, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8851516246795654, "num_tokens": 376722836.0, "step": 9873 }, { "epoch": 1.256074290802697, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.047908782958984, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8606398105621338, "num_tokens": 376763194.0, "step": 9874 }, { "epoch": 1.2562015010812875, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.969515800476074, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8627666234970093, "num_tokens": 376803701.0, "step": 9875 }, { "epoch": 1.256328711359878, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.000682830810547, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8762292861938477, "num_tokens": 376835296.0, "step": 9876 }, { "epoch": 1.2564559216384683, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.023414611816406, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8711899518966675, "num_tokens": 376876317.0, "step": 9877 }, { "epoch": 1.2565831319170588, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.96985149383545, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.865307629108429, "num_tokens": 376916403.0, "step": 9878 }, { "epoch": 1.2567103421956494, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.045286178588867, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8765109777450562, "num_tokens": 376956079.0, "step": 9879 }, { "epoch": 1.25683755247424, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.035446166992188, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8576949834823608, "num_tokens": 376991618.0, "step": 9880 }, { "epoch": 1.2569647627528304, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.010887145996094, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8737711906433105, "num_tokens": 377033964.0, "step": 9881 }, { "epoch": 1.257091973031421, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.997696876525879, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8747339248657227, "num_tokens": 377070091.0, "step": 9882 }, { "epoch": 1.2572191833100115, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.995288848876953, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8752315640449524, "num_tokens": 377111185.0, "step": 9883 }, { "epoch": 1.257346393588602, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.026632308959961, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8727880716323853, "num_tokens": 377147637.0, "step": 9884 }, { "epoch": 1.2574736038671925, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.025908470153809, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.867219090461731, "num_tokens": 377181918.0, "step": 9885 }, { "epoch": 1.2576008141457828, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.040797233581543, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8621405363082886, "num_tokens": 377222900.0, "step": 9886 }, { "epoch": 1.2577280244243734, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.031827926635742, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8818799257278442, "num_tokens": 377263107.0, "step": 9887 }, { "epoch": 1.257855234702964, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.944971084594727, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.88286292552948, "num_tokens": 377301384.0, "step": 9888 }, { "epoch": 1.2579824449815544, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.041031837463379, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.861456573009491, "num_tokens": 377339865.0, "step": 9889 }, { "epoch": 1.258109655260145, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.99057388305664, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8766957521438599, "num_tokens": 377381582.0, "step": 9890 }, { "epoch": 1.2582368655387355, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.04738998413086, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8445762395858765, "num_tokens": 377425548.0, "step": 9891 }, { "epoch": 1.258364075817326, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.94992733001709, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8669165372848511, "num_tokens": 377462298.0, "step": 9892 }, { "epoch": 1.2584912860959165, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.087179183959961, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8597862124443054, "num_tokens": 377503873.0, "step": 9893 }, { "epoch": 1.258618496374507, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.999322891235352, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8822511434555054, "num_tokens": 377540497.0, "step": 9894 }, { "epoch": 1.2587457066530976, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.016227722167969, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8681633472442627, "num_tokens": 377582242.0, "step": 9895 }, { "epoch": 1.2588729169316881, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.971819877624512, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8831503391265869, "num_tokens": 377614972.0, "step": 9896 }, { "epoch": 1.2590001272102787, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.082547187805176, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8646553158760071, "num_tokens": 377645392.0, "step": 9897 }, { "epoch": 1.2591273374888692, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.969478607177734, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8757346868515015, "num_tokens": 377680317.0, "step": 9898 }, { "epoch": 1.2592545477674597, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.065957069396973, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8673333525657654, "num_tokens": 377714879.0, "step": 9899 }, { "epoch": 1.2593817580460502, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.982117652893066, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8528609275817871, "num_tokens": 377751404.0, "step": 9900 }, { "epoch": 1.2595089683246408, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.036154747009277, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8713289499282837, "num_tokens": 377788354.0, "step": 9901 }, { "epoch": 1.259636178603231, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.996392250061035, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8579450845718384, "num_tokens": 377827259.0, "step": 9902 }, { "epoch": 1.2597633888818216, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.071136474609375, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.85963374376297, "num_tokens": 377872477.0, "step": 9903 }, { "epoch": 1.2598905991604121, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.038060188293457, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8606589436531067, "num_tokens": 377911159.0, "step": 9904 }, { "epoch": 1.2600178094390027, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.080116271972656, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8714200258255005, "num_tokens": 377951919.0, "step": 9905 }, { "epoch": 1.2601450197175932, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.038019180297852, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8493501543998718, "num_tokens": 377997567.0, "step": 9906 }, { "epoch": 1.2602722299961837, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.043264389038086, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8662766218185425, "num_tokens": 378033811.0, "step": 9907 }, { "epoch": 1.2603994402747742, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.057024955749512, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8549536466598511, "num_tokens": 378070136.0, "step": 9908 }, { "epoch": 1.2605266505533648, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.117452621459961, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8648855090141296, "num_tokens": 378106902.0, "step": 9909 }, { "epoch": 1.2606538608319553, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.965435981750488, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.871181845664978, "num_tokens": 378144638.0, "step": 9910 }, { "epoch": 1.2607810711105456, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.102519035339355, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8769494295120239, "num_tokens": 378179059.0, "step": 9911 }, { "epoch": 1.2609082813891361, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.097426414489746, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8554227352142334, "num_tokens": 378210346.0, "step": 9912 }, { "epoch": 1.2610354916677267, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.023774147033691, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.873245894908905, "num_tokens": 378242687.0, "step": 9913 }, { "epoch": 1.2611627019463172, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.068849563598633, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8733094334602356, "num_tokens": 378279421.0, "step": 9914 }, { "epoch": 1.2612899122249077, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.963510513305664, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8629765510559082, "num_tokens": 378318133.0, "step": 9915 }, { "epoch": 1.2614171225034982, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.042930603027344, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8496854305267334, "num_tokens": 378362987.0, "step": 9916 }, { "epoch": 1.2615443327820888, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.038878440856934, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8842628598213196, "num_tokens": 378396197.0, "step": 9917 }, { "epoch": 1.2616715430606793, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.024173736572266, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8757503628730774, "num_tokens": 378433691.0, "step": 9918 }, { "epoch": 1.2617987533392698, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.036579132080078, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8709166646003723, "num_tokens": 378470277.0, "step": 9919 }, { "epoch": 1.2619259636178604, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.14250373840332, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.861272931098938, "num_tokens": 378510500.0, "step": 9920 }, { "epoch": 1.2620531738964509, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.990127563476562, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8654385209083557, "num_tokens": 378552893.0, "step": 9921 }, { "epoch": 1.2621803841750414, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.149948120117188, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8740143775939941, "num_tokens": 378590917.0, "step": 9922 }, { "epoch": 1.262307594453632, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.007817268371582, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.874874472618103, "num_tokens": 378637764.0, "step": 9923 }, { "epoch": 1.2624348047322225, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.123908996582031, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8639029860496521, "num_tokens": 378675497.0, "step": 9924 }, { "epoch": 1.262562015010813, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.087311744689941, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8584568500518799, "num_tokens": 378710470.0, "step": 9925 }, { "epoch": 1.2626892252894033, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.06346321105957, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8635366559028625, "num_tokens": 378758116.0, "step": 9926 }, { "epoch": 1.2628164355679938, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.041863441467285, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8646790385246277, "num_tokens": 378798166.0, "step": 9927 }, { "epoch": 1.2629436458465844, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.112771987915039, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8824452757835388, "num_tokens": 378839899.0, "step": 9928 }, { "epoch": 1.263070856125175, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.02060604095459, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8686120510101318, "num_tokens": 378886022.0, "step": 9929 }, { "epoch": 1.2631980664037654, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.053895950317383, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8569554686546326, "num_tokens": 378925097.0, "step": 9930 }, { "epoch": 1.263325276682356, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.978475570678711, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8794524669647217, "num_tokens": 378965242.0, "step": 9931 }, { "epoch": 1.2634524869609465, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.057531356811523, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.8473122715950012, "num_tokens": 379005932.0, "step": 9932 }, { "epoch": 1.263579697239537, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.989191055297852, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8650652170181274, "num_tokens": 379045817.0, "step": 9933 }, { "epoch": 1.2637069075181275, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.065485000610352, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8603596687316895, "num_tokens": 379086481.0, "step": 9934 }, { "epoch": 1.2638341177967178, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.985838890075684, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8504371643066406, "num_tokens": 379129782.0, "step": 9935 }, { "epoch": 1.2639613280753084, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.013294219970703, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8648557662963867, "num_tokens": 379171525.0, "step": 9936 }, { "epoch": 1.264088538353899, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.08466625213623, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.860787034034729, "num_tokens": 379213887.0, "step": 9937 }, { "epoch": 1.2642157486324894, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.040393829345703, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8673790693283081, "num_tokens": 379252634.0, "step": 9938 }, { "epoch": 1.26434295891108, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.048598289489746, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8665562868118286, "num_tokens": 379287402.0, "step": 9939 }, { "epoch": 1.2644701691896705, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 12.100481033325195, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8727774620056152, "num_tokens": 379325592.0, "step": 9940 }, { "epoch": 1.264597379468261, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.040645599365234, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8551148176193237, "num_tokens": 379359010.0, "step": 9941 }, { "epoch": 1.2647245897468515, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.111686706542969, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8597854375839233, "num_tokens": 379399783.0, "step": 9942 }, { "epoch": 1.264851800025442, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.041208267211914, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8705882430076599, "num_tokens": 379442207.0, "step": 9943 }, { "epoch": 1.2649790103040326, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.055078506469727, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.866587221622467, "num_tokens": 379479561.0, "step": 9944 }, { "epoch": 1.2651062205826231, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.10083293914795, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8727285265922546, "num_tokens": 379516413.0, "step": 9945 }, { "epoch": 1.2652334308612136, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.007364273071289, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8581183552742004, "num_tokens": 379557993.0, "step": 9946 }, { "epoch": 1.2653606411398042, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.058541297912598, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8489687442779541, "num_tokens": 379594028.0, "step": 9947 }, { "epoch": 1.2654878514183947, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.094796180725098, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8613700866699219, "num_tokens": 379631235.0, "step": 9948 }, { "epoch": 1.2656150616969852, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.029288291931152, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8559526801109314, "num_tokens": 379668714.0, "step": 9949 }, { "epoch": 1.2657422719755758, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.068282127380371, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8721417188644409, "num_tokens": 379710233.0, "step": 9950 }, { "epoch": 1.265869482254166, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.079215049743652, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8643660545349121, "num_tokens": 379746208.0, "step": 9951 }, { "epoch": 1.2659966925327566, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.989548683166504, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8766310811042786, "num_tokens": 379796420.0, "step": 9952 }, { "epoch": 1.2661239028113471, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.074919700622559, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.868620753288269, "num_tokens": 379830027.0, "step": 9953 }, { "epoch": 1.2662511130899377, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.053423881530762, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8673694133758545, "num_tokens": 379871751.0, "step": 9954 }, { "epoch": 1.2663783233685282, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.06088638305664, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8600432276725769, "num_tokens": 379908255.0, "step": 9955 }, { "epoch": 1.2665055336471187, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.098623275756836, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8699206709861755, "num_tokens": 379944502.0, "step": 9956 }, { "epoch": 1.2666327439257092, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.052713394165039, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8641775846481323, "num_tokens": 379981929.0, "step": 9957 }, { "epoch": 1.2667599542042998, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.082079887390137, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8615654706954956, "num_tokens": 380018876.0, "step": 9958 }, { "epoch": 1.2668871644828903, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.026578903198242, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8613438606262207, "num_tokens": 380054153.0, "step": 9959 }, { "epoch": 1.2670143747614806, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.130605697631836, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8582842946052551, "num_tokens": 380086550.0, "step": 9960 }, { "epoch": 1.2671415850400711, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.036174774169922, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8613787889480591, "num_tokens": 380120308.0, "step": 9961 }, { "epoch": 1.2672687953186617, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.076614379882812, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8502366542816162, "num_tokens": 380160204.0, "step": 9962 }, { "epoch": 1.2673960055972522, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.044958114624023, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.870002031326294, "num_tokens": 380196687.0, "step": 9963 }, { "epoch": 1.2675232158758427, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.07931137084961, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8814929723739624, "num_tokens": 380237451.0, "step": 9964 }, { "epoch": 1.2676504261544332, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.001616477966309, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8620415329933167, "num_tokens": 380276631.0, "step": 9965 }, { "epoch": 1.2677776364330238, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.021961212158203, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.871712863445282, "num_tokens": 380319432.0, "step": 9966 }, { "epoch": 1.2679048467116143, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.037541389465332, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8743029832839966, "num_tokens": 380356666.0, "step": 9967 }, { "epoch": 1.2680320569902048, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.031488418579102, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8748277425765991, "num_tokens": 380390347.0, "step": 9968 }, { "epoch": 1.2681592672687954, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.059747695922852, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8706724643707275, "num_tokens": 380424083.0, "step": 9969 }, { "epoch": 1.2682864775473859, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.028681755065918, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8609790802001953, "num_tokens": 380471179.0, "step": 9970 }, { "epoch": 1.2684136878259764, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.075912475585938, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.868202805519104, "num_tokens": 380510291.0, "step": 9971 }, { "epoch": 1.268540898104567, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.021038055419922, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8715629577636719, "num_tokens": 380556297.0, "step": 9972 }, { "epoch": 1.2686681083831575, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.07143497467041, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8769770860671997, "num_tokens": 380592113.0, "step": 9973 }, { "epoch": 1.268795318661748, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.092223167419434, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8799500465393066, "num_tokens": 380623942.0, "step": 9974 }, { "epoch": 1.2689225289403383, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.002260208129883, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8603942394256592, "num_tokens": 380662300.0, "step": 9975 }, { "epoch": 1.2690497392189288, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.051539421081543, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8687448501586914, "num_tokens": 380701452.0, "step": 9976 }, { "epoch": 1.2691769494975194, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.01668930053711, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8774559497833252, "num_tokens": 380741248.0, "step": 9977 }, { "epoch": 1.2693041597761099, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.061723709106445, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8604618310928345, "num_tokens": 380781455.0, "step": 9978 }, { "epoch": 1.2694313700547004, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.075626373291016, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8588491678237915, "num_tokens": 380818139.0, "step": 9979 }, { "epoch": 1.269558580333291, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.956551551818848, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8727020621299744, "num_tokens": 380860530.0, "step": 9980 }, { "epoch": 1.2696857906118815, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.070602416992188, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8787491321563721, "num_tokens": 380900825.0, "step": 9981 }, { "epoch": 1.269813000890472, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.010526657104492, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.864495038986206, "num_tokens": 380939270.0, "step": 9982 }, { "epoch": 1.2699402111690625, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.029216766357422, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.8442784547805786, "num_tokens": 380981708.0, "step": 9983 }, { "epoch": 1.2700674214476528, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.07972526550293, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8728318810462952, "num_tokens": 381015186.0, "step": 9984 }, { "epoch": 1.2701946317262434, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.091145515441895, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8772061467170715, "num_tokens": 381055503.0, "step": 9985 }, { "epoch": 1.270321842004834, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.0219087600708, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8796859979629517, "num_tokens": 381091935.0, "step": 9986 }, { "epoch": 1.2704490522834244, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.072086334228516, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8717505931854248, "num_tokens": 381127951.0, "step": 9987 }, { "epoch": 1.270576262562015, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.043604850769043, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8618260622024536, "num_tokens": 381161918.0, "step": 9988 }, { "epoch": 1.2707034728406055, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.100397109985352, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8556628823280334, "num_tokens": 381198747.0, "step": 9989 }, { "epoch": 1.270830683119196, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.057043075561523, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8682370185852051, "num_tokens": 381242714.0, "step": 9990 }, { "epoch": 1.2709578933977865, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.062276840209961, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8793525099754333, "num_tokens": 381274667.0, "step": 9991 }, { "epoch": 1.271085103676377, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.075654029846191, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8772932291030884, "num_tokens": 381318110.0, "step": 9992 }, { "epoch": 1.2712123139549676, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.048016548156738, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8679105043411255, "num_tokens": 381356694.0, "step": 9993 }, { "epoch": 1.2713395242335581, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.084617614746094, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8572645783424377, "num_tokens": 381395218.0, "step": 9994 }, { "epoch": 1.2714667345121486, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.14189338684082, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8803094625473022, "num_tokens": 381429195.0, "step": 9995 }, { "epoch": 1.2715939447907392, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.052247047424316, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8617130517959595, "num_tokens": 381468276.0, "step": 9996 }, { "epoch": 1.2717211550693297, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.135354042053223, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8756580352783203, "num_tokens": 381503400.0, "step": 9997 }, { "epoch": 1.2718483653479202, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.057296752929688, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.873021125793457, "num_tokens": 381542126.0, "step": 9998 }, { "epoch": 1.2719755756265108, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.034092903137207, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8764264583587646, "num_tokens": 381579169.0, "step": 9999 }, { "epoch": 1.272102785905101, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.125738143920898, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8531180620193481, "num_tokens": 381615501.0, "step": 10000 }, { "epoch": 1.2722299961836916, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.097854614257812, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8625009059906006, "num_tokens": 381649737.0, "step": 10001 }, { "epoch": 1.2723572064622821, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.041585922241211, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8711278438568115, "num_tokens": 381693776.0, "step": 10002 }, { "epoch": 1.2724844167408726, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.074389457702637, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8772533535957336, "num_tokens": 381733324.0, "step": 10003 }, { "epoch": 1.2726116270194632, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.051609992980957, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.863456130027771, "num_tokens": 381765893.0, "step": 10004 }, { "epoch": 1.2727388372980537, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.038932800292969, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8652315139770508, "num_tokens": 381802782.0, "step": 10005 }, { "epoch": 1.2728660475766442, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.096487045288086, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8701251745223999, "num_tokens": 381843448.0, "step": 10006 }, { "epoch": 1.2729932578552348, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.056994438171387, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8569236993789673, "num_tokens": 381878850.0, "step": 10007 }, { "epoch": 1.2731204681338253, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.073467254638672, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8466839790344238, "num_tokens": 381915508.0, "step": 10008 }, { "epoch": 1.2732476784124156, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.066035270690918, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.857434868812561, "num_tokens": 381952699.0, "step": 10009 }, { "epoch": 1.2733748886910061, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.057672500610352, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8608168363571167, "num_tokens": 381989947.0, "step": 10010 }, { "epoch": 1.2735020989695967, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.068757057189941, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8669065237045288, "num_tokens": 382027701.0, "step": 10011 }, { "epoch": 1.2736293092481872, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.013973236083984, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8652675747871399, "num_tokens": 382066415.0, "step": 10012 }, { "epoch": 1.2737565195267777, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.099177360534668, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8622533679008484, "num_tokens": 382108402.0, "step": 10013 }, { "epoch": 1.2738837298053682, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.12324333190918, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8670163154602051, "num_tokens": 382146865.0, "step": 10014 }, { "epoch": 1.2740109400839588, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.010212898254395, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8646621704101562, "num_tokens": 382186746.0, "step": 10015 }, { "epoch": 1.2741381503625493, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.096713066101074, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8807661533355713, "num_tokens": 382220693.0, "step": 10016 }, { "epoch": 1.2742653606411398, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.025753021240234, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8611318469047546, "num_tokens": 382259549.0, "step": 10017 }, { "epoch": 1.2743925709197303, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.054643630981445, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8774278163909912, "num_tokens": 382299164.0, "step": 10018 }, { "epoch": 1.2745197811983209, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.047477722167969, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8802226781845093, "num_tokens": 382333596.0, "step": 10019 }, { "epoch": 1.2746469914769114, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.0719575881958, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8684607148170471, "num_tokens": 382372321.0, "step": 10020 }, { "epoch": 1.274774201755502, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.031789779663086, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8693352937698364, "num_tokens": 382411125.0, "step": 10021 }, { "epoch": 1.2749014120340925, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.09455394744873, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8571136593818665, "num_tokens": 382450935.0, "step": 10022 }, { "epoch": 1.275028622312683, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.113122940063477, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8656504154205322, "num_tokens": 382488232.0, "step": 10023 }, { "epoch": 1.2751558325912733, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.037863731384277, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8630244731903076, "num_tokens": 382522280.0, "step": 10024 }, { "epoch": 1.2752830428698638, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.990503311157227, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8771881461143494, "num_tokens": 382562788.0, "step": 10025 }, { "epoch": 1.2754102531484544, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.060434341430664, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8600123524665833, "num_tokens": 382598534.0, "step": 10026 }, { "epoch": 1.2755374634270449, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.035357475280762, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8610373735427856, "num_tokens": 382641803.0, "step": 10027 }, { "epoch": 1.2756646737056354, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.0512113571167, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8785349130630493, "num_tokens": 382677871.0, "step": 10028 }, { "epoch": 1.275791883984226, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.007234573364258, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.876184344291687, "num_tokens": 382714989.0, "step": 10029 }, { "epoch": 1.2759190942628165, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.100661277770996, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8704711198806763, "num_tokens": 382751919.0, "step": 10030 }, { "epoch": 1.276046304541407, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.958864212036133, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8784394264221191, "num_tokens": 382793550.0, "step": 10031 }, { "epoch": 1.2761735148199975, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.079142570495605, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8745245337486267, "num_tokens": 382833142.0, "step": 10032 }, { "epoch": 1.2763007250985878, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.10327434539795, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8751536011695862, "num_tokens": 382869934.0, "step": 10033 }, { "epoch": 1.2764279353771784, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.034693717956543, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8544808030128479, "num_tokens": 382909533.0, "step": 10034 }, { "epoch": 1.2765551456557689, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.091318130493164, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8728077411651611, "num_tokens": 382947462.0, "step": 10035 }, { "epoch": 1.2766823559343594, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.08502197265625, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8688697218894958, "num_tokens": 382989973.0, "step": 10036 }, { "epoch": 1.27680956621295, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.98098373413086, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8653736114501953, "num_tokens": 383033052.0, "step": 10037 }, { "epoch": 1.2769367764915405, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.060023307800293, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8657792806625366, "num_tokens": 383075981.0, "step": 10038 }, { "epoch": 1.277063986770131, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.04244327545166, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8840424418449402, "num_tokens": 383111984.0, "step": 10039 }, { "epoch": 1.2771911970487215, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.010518074035645, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8679195642471313, "num_tokens": 383155323.0, "step": 10040 }, { "epoch": 1.277318407327312, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.090166091918945, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8662143349647522, "num_tokens": 383190069.0, "step": 10041 }, { "epoch": 1.2774456176059026, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.00851058959961, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8596286773681641, "num_tokens": 383224999.0, "step": 10042 }, { "epoch": 1.2775728278844931, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.078388214111328, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8598687052726746, "num_tokens": 383267764.0, "step": 10043 }, { "epoch": 1.2777000381630836, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.064878463745117, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8768300414085388, "num_tokens": 383306148.0, "step": 10044 }, { "epoch": 1.2778272484416742, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.06587028503418, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8718633651733398, "num_tokens": 383343408.0, "step": 10045 }, { "epoch": 1.2779544587202647, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.054380416870117, "learning_rate": 1e-06, "loss": 0.334, "mean_token_accuracy": 0.8912547826766968, "num_tokens": 383383626.0, "step": 10046 }, { "epoch": 1.2780816689988552, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.12996768951416, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8514803051948547, "num_tokens": 383423218.0, "step": 10047 }, { "epoch": 1.2782088792774458, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.069960594177246, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8745871782302856, "num_tokens": 383461126.0, "step": 10048 }, { "epoch": 1.278336089556036, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.003158569335938, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8594549894332886, "num_tokens": 383501808.0, "step": 10049 }, { "epoch": 1.2784632998346266, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.077312469482422, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8878054618835449, "num_tokens": 383542395.0, "step": 10050 }, { "epoch": 1.2785905101132171, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.059235572814941, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8642141819000244, "num_tokens": 383579968.0, "step": 10051 }, { "epoch": 1.2787177203918076, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.016507148742676, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8671612739562988, "num_tokens": 383617861.0, "step": 10052 }, { "epoch": 1.2788449306703982, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.145124435424805, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8631879687309265, "num_tokens": 383652534.0, "step": 10053 }, { "epoch": 1.2789721409489887, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.03753662109375, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8626803159713745, "num_tokens": 383693120.0, "step": 10054 }, { "epoch": 1.2790993512275792, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.092382431030273, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.858771562576294, "num_tokens": 383734958.0, "step": 10055 }, { "epoch": 1.2792265615061698, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.1181640625, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.860934853553772, "num_tokens": 383768873.0, "step": 10056 }, { "epoch": 1.2793537717847603, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.034070014953613, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8690677881240845, "num_tokens": 383804063.0, "step": 10057 }, { "epoch": 1.2794809820633506, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.017863273620605, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8815306425094604, "num_tokens": 383843098.0, "step": 10058 }, { "epoch": 1.2796081923419411, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.086579322814941, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8761391639709473, "num_tokens": 383879443.0, "step": 10059 }, { "epoch": 1.2797354026205316, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.046613693237305, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8672177195549011, "num_tokens": 383918071.0, "step": 10060 }, { "epoch": 1.2798626128991222, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.022501945495605, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8615334033966064, "num_tokens": 383953052.0, "step": 10061 }, { "epoch": 1.2799898231777127, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.079141616821289, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8653579950332642, "num_tokens": 383992835.0, "step": 10062 }, { "epoch": 1.2801170334563032, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.05535888671875, "learning_rate": 1e-06, "loss": 0.4735, "mean_token_accuracy": 0.8546796441078186, "num_tokens": 384031234.0, "step": 10063 }, { "epoch": 1.2802442437348938, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.1046781539917, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8596320152282715, "num_tokens": 384063611.0, "step": 10064 }, { "epoch": 1.2803714540134843, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.005881309509277, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8762952089309692, "num_tokens": 384106757.0, "step": 10065 }, { "epoch": 1.2804986642920748, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.066268920898438, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8714772462844849, "num_tokens": 384143094.0, "step": 10066 }, { "epoch": 1.2806258745706653, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.038017272949219, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8673940896987915, "num_tokens": 384181498.0, "step": 10067 }, { "epoch": 1.2807530848492559, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.145112037658691, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.871814489364624, "num_tokens": 384209784.0, "step": 10068 }, { "epoch": 1.2808802951278464, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.08230209350586, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8733541369438171, "num_tokens": 384248000.0, "step": 10069 }, { "epoch": 1.281007505406437, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.045025825500488, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8585054278373718, "num_tokens": 384290897.0, "step": 10070 }, { "epoch": 1.2811347156850275, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.053638458251953, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8610307574272156, "num_tokens": 384329464.0, "step": 10071 }, { "epoch": 1.281261925963618, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.030482292175293, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8709815740585327, "num_tokens": 384368915.0, "step": 10072 }, { "epoch": 1.2813891362422083, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.069948196411133, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8765022158622742, "num_tokens": 384407445.0, "step": 10073 }, { "epoch": 1.2815163465207988, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.019147872924805, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8634929060935974, "num_tokens": 384444384.0, "step": 10074 }, { "epoch": 1.2816435567993893, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.129254341125488, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8649582862854004, "num_tokens": 384479670.0, "step": 10075 }, { "epoch": 1.2817707670779799, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.990188598632812, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8562678694725037, "num_tokens": 384522436.0, "step": 10076 }, { "epoch": 1.2818979773565704, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.053472518920898, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.860392153263092, "num_tokens": 384559754.0, "step": 10077 }, { "epoch": 1.282025187635161, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.130996704101562, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8644227981567383, "num_tokens": 384595153.0, "step": 10078 }, { "epoch": 1.2821523979137515, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.06912899017334, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8667305707931519, "num_tokens": 384628345.0, "step": 10079 }, { "epoch": 1.282279608192342, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.119097709655762, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8627345561981201, "num_tokens": 384666725.0, "step": 10080 }, { "epoch": 1.2824068184709325, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.03756332397461, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8582308888435364, "num_tokens": 384706217.0, "step": 10081 }, { "epoch": 1.2825340287495228, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.154714584350586, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8640069961547852, "num_tokens": 384741835.0, "step": 10082 }, { "epoch": 1.2826612390281134, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.045004844665527, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8684012293815613, "num_tokens": 384777273.0, "step": 10083 }, { "epoch": 1.2827884493067039, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.059613227844238, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8810433745384216, "num_tokens": 384814950.0, "step": 10084 }, { "epoch": 1.2829156595852944, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.111265182495117, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8595374822616577, "num_tokens": 384848088.0, "step": 10085 }, { "epoch": 1.283042869863885, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.07537841796875, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8593258857727051, "num_tokens": 384887361.0, "step": 10086 }, { "epoch": 1.2831700801424755, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.032246589660645, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8731536865234375, "num_tokens": 384930459.0, "step": 10087 }, { "epoch": 1.283297290421066, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.045578956604004, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8614374399185181, "num_tokens": 384969506.0, "step": 10088 }, { "epoch": 1.2834245006996565, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.086346626281738, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8734208941459656, "num_tokens": 385004368.0, "step": 10089 }, { "epoch": 1.283551710978247, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.108267784118652, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8690937161445618, "num_tokens": 385044553.0, "step": 10090 }, { "epoch": 1.2836789212568376, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.070076942443848, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8794295787811279, "num_tokens": 385076873.0, "step": 10091 }, { "epoch": 1.283806131535428, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.019051551818848, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8700573444366455, "num_tokens": 385113969.0, "step": 10092 }, { "epoch": 1.2839333418140186, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.083223342895508, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8770056962966919, "num_tokens": 385149460.0, "step": 10093 }, { "epoch": 1.2840605520926092, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.040377616882324, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8506090044975281, "num_tokens": 385187240.0, "step": 10094 }, { "epoch": 1.2841877623711997, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.084568977355957, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8594790697097778, "num_tokens": 385225532.0, "step": 10095 }, { "epoch": 1.2843149726497902, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.066102981567383, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8698972463607788, "num_tokens": 385267404.0, "step": 10096 }, { "epoch": 1.2844421829283807, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.96767520904541, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8559818267822266, "num_tokens": 385314008.0, "step": 10097 }, { "epoch": 1.284569393206971, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.11849594116211, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.854986310005188, "num_tokens": 385352204.0, "step": 10098 }, { "epoch": 1.2846966034855616, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.015091896057129, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.858893632888794, "num_tokens": 385393307.0, "step": 10099 }, { "epoch": 1.2848238137641521, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.089856147766113, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8680554628372192, "num_tokens": 385426124.0, "step": 10100 }, { "epoch": 1.2849510240427426, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.029598236083984, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8727387189865112, "num_tokens": 385467471.0, "step": 10101 }, { "epoch": 1.2850782343213332, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.001578330993652, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8581401109695435, "num_tokens": 385508780.0, "step": 10102 }, { "epoch": 1.2852054445999237, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.023032188415527, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8725391626358032, "num_tokens": 385552646.0, "step": 10103 }, { "epoch": 1.2853326548785142, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.117597579956055, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8546292781829834, "num_tokens": 385586676.0, "step": 10104 }, { "epoch": 1.2854598651571048, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.059765815734863, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8558157086372375, "num_tokens": 385623201.0, "step": 10105 }, { "epoch": 1.2855870754356953, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.072681427001953, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8780866861343384, "num_tokens": 385662716.0, "step": 10106 }, { "epoch": 1.2857142857142856, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.047269821166992, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.856701672077179, "num_tokens": 385706986.0, "step": 10107 }, { "epoch": 1.2858414959928761, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.081413269042969, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8664172887802124, "num_tokens": 385744240.0, "step": 10108 }, { "epoch": 1.2859687062714666, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.047953605651855, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8509393930435181, "num_tokens": 385778583.0, "step": 10109 }, { "epoch": 1.2860959165500572, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.043944358825684, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8674319982528687, "num_tokens": 385819910.0, "step": 10110 }, { "epoch": 1.2862231268286477, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.077301025390625, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8645976185798645, "num_tokens": 385860068.0, "step": 10111 }, { "epoch": 1.2863503371072382, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.107550621032715, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8592872619628906, "num_tokens": 385890460.0, "step": 10112 }, { "epoch": 1.2864775473858288, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.096921920776367, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8638415336608887, "num_tokens": 385926174.0, "step": 10113 }, { "epoch": 1.2866047576644193, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.995842933654785, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8765166997909546, "num_tokens": 385962928.0, "step": 10114 }, { "epoch": 1.2867319679430098, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.050886154174805, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8748898506164551, "num_tokens": 386002509.0, "step": 10115 }, { "epoch": 1.2868591782216003, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.985447883605957, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8668916821479797, "num_tokens": 386043304.0, "step": 10116 }, { "epoch": 1.2869863885001909, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.036002159118652, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8739080429077148, "num_tokens": 386081607.0, "step": 10117 }, { "epoch": 1.2871135987787814, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.01953125, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.88524329662323, "num_tokens": 386117762.0, "step": 10118 }, { "epoch": 1.287240809057372, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.022316932678223, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8590966463088989, "num_tokens": 386155908.0, "step": 10119 }, { "epoch": 1.2873680193359625, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.003466606140137, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8645982146263123, "num_tokens": 386199385.0, "step": 10120 }, { "epoch": 1.287495229614553, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.14865493774414, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.862561821937561, "num_tokens": 386231892.0, "step": 10121 }, { "epoch": 1.2876224398931433, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.06354808807373, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8513819575309753, "num_tokens": 386269657.0, "step": 10122 }, { "epoch": 1.2877496501717338, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.984552383422852, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8714861273765564, "num_tokens": 386306936.0, "step": 10123 }, { "epoch": 1.2878768604503243, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.066290855407715, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8673014044761658, "num_tokens": 386342260.0, "step": 10124 }, { "epoch": 1.2880040707289149, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.050214767456055, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8624894618988037, "num_tokens": 386381784.0, "step": 10125 }, { "epoch": 1.2881312810075054, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.124358177185059, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8662590980529785, "num_tokens": 386422350.0, "step": 10126 }, { "epoch": 1.288258491286096, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.05653190612793, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8631601333618164, "num_tokens": 386467951.0, "step": 10127 }, { "epoch": 1.2883857015646865, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.030404090881348, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8755593299865723, "num_tokens": 386511896.0, "step": 10128 }, { "epoch": 1.288512911843277, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.072324752807617, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8542245030403137, "num_tokens": 386550704.0, "step": 10129 }, { "epoch": 1.2886401221218675, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.098854064941406, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8502030968666077, "num_tokens": 386589295.0, "step": 10130 }, { "epoch": 1.2887673324004578, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.076422691345215, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8661025166511536, "num_tokens": 386623462.0, "step": 10131 }, { "epoch": 1.2888945426790483, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.069299697875977, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8696916103363037, "num_tokens": 386663036.0, "step": 10132 }, { "epoch": 1.2890217529576389, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.06567668914795, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.866690993309021, "num_tokens": 386702530.0, "step": 10133 }, { "epoch": 1.2891489632362294, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.094218254089355, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8813953399658203, "num_tokens": 386739758.0, "step": 10134 }, { "epoch": 1.28927617351482, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.046381950378418, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8768035173416138, "num_tokens": 386778287.0, "step": 10135 }, { "epoch": 1.2894033837934105, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.055994033813477, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8771612644195557, "num_tokens": 386812348.0, "step": 10136 }, { "epoch": 1.289530594072001, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.030415534973145, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8736528754234314, "num_tokens": 386855540.0, "step": 10137 }, { "epoch": 1.2896578043505915, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.05946159362793, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8622589111328125, "num_tokens": 386893775.0, "step": 10138 }, { "epoch": 1.289785014629182, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.027168273925781, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8675149083137512, "num_tokens": 386929553.0, "step": 10139 }, { "epoch": 1.2899122249077726, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.05929946899414, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8617106080055237, "num_tokens": 386968033.0, "step": 10140 }, { "epoch": 1.290039435186363, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.065133094787598, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8655242323875427, "num_tokens": 387004967.0, "step": 10141 }, { "epoch": 1.2901666454649536, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.106222152709961, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8554015755653381, "num_tokens": 387038086.0, "step": 10142 }, { "epoch": 1.2902938557435442, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.056990623474121, "learning_rate": 1e-06, "loss": 0.4988, "mean_token_accuracy": 0.8414918780326843, "num_tokens": 387077112.0, "step": 10143 }, { "epoch": 1.2904210660221347, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.085155487060547, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8503381013870239, "num_tokens": 387118800.0, "step": 10144 }, { "epoch": 1.2905482763007252, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.076064109802246, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8610140681266785, "num_tokens": 387160321.0, "step": 10145 }, { "epoch": 1.2906754865793157, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.109403610229492, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.864154577255249, "num_tokens": 387199387.0, "step": 10146 }, { "epoch": 1.290802696857906, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.016889572143555, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.873133659362793, "num_tokens": 387238859.0, "step": 10147 }, { "epoch": 1.2909299071364966, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.108222007751465, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8724830746650696, "num_tokens": 387272149.0, "step": 10148 }, { "epoch": 1.291057117415087, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.04002571105957, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8657061457633972, "num_tokens": 387311617.0, "step": 10149 }, { "epoch": 1.2911843276936776, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.103854179382324, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8536503911018372, "num_tokens": 387344281.0, "step": 10150 }, { "epoch": 1.2913115379722682, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.083600044250488, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8562010526657104, "num_tokens": 387385959.0, "step": 10151 }, { "epoch": 1.2914387482508587, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.107656478881836, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8616842031478882, "num_tokens": 387426197.0, "step": 10152 }, { "epoch": 1.2915659585294492, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.106782913208008, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8727253079414368, "num_tokens": 387462250.0, "step": 10153 }, { "epoch": 1.2916931688080397, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.995403289794922, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8659759759902954, "num_tokens": 387499903.0, "step": 10154 }, { "epoch": 1.2918203790866303, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.091065406799316, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8728117346763611, "num_tokens": 387535361.0, "step": 10155 }, { "epoch": 1.2919475893652206, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.050444602966309, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8627446889877319, "num_tokens": 387574300.0, "step": 10156 }, { "epoch": 1.2920747996438111, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.112351417541504, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8690273761749268, "num_tokens": 387606860.0, "step": 10157 }, { "epoch": 1.2922020099224016, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.02237606048584, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8564849495887756, "num_tokens": 387646519.0, "step": 10158 }, { "epoch": 1.2923292202009922, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.090208053588867, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.878304123878479, "num_tokens": 387681100.0, "step": 10159 }, { "epoch": 1.2924564304795827, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.136258125305176, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8884181380271912, "num_tokens": 387715962.0, "step": 10160 }, { "epoch": 1.2925836407581732, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.055474281311035, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8651711940765381, "num_tokens": 387755856.0, "step": 10161 }, { "epoch": 1.2927108510367638, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.131781578063965, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8749801516532898, "num_tokens": 387793256.0, "step": 10162 }, { "epoch": 1.2928380613153543, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.04554557800293, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8628071546554565, "num_tokens": 387834696.0, "step": 10163 }, { "epoch": 1.2929652715939448, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.077642440795898, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8552320003509521, "num_tokens": 387874973.0, "step": 10164 }, { "epoch": 1.2930924818725353, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.07911205291748, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8564603924751282, "num_tokens": 387904887.0, "step": 10165 }, { "epoch": 1.2932196921511259, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.087472915649414, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8742332458496094, "num_tokens": 387944676.0, "step": 10166 }, { "epoch": 1.2933469024297164, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.10004711151123, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8673798441886902, "num_tokens": 387982234.0, "step": 10167 }, { "epoch": 1.293474112708307, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.049348831176758, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8689660429954529, "num_tokens": 388018001.0, "step": 10168 }, { "epoch": 1.2936013229868975, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.088602066040039, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8617373704910278, "num_tokens": 388057476.0, "step": 10169 }, { "epoch": 1.293728533265488, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.072875022888184, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8843082189559937, "num_tokens": 388095351.0, "step": 10170 }, { "epoch": 1.2938557435440783, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.163984298706055, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8661390542984009, "num_tokens": 388135366.0, "step": 10171 }, { "epoch": 1.2939829538226688, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.083104133605957, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8559641242027283, "num_tokens": 388175003.0, "step": 10172 }, { "epoch": 1.2941101641012593, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.1391019821167, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8477463722229004, "num_tokens": 388207083.0, "step": 10173 }, { "epoch": 1.2942373743798499, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.141363143920898, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8821779489517212, "num_tokens": 388245038.0, "step": 10174 }, { "epoch": 1.2943645846584404, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.115602493286133, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.862481415271759, "num_tokens": 388285035.0, "step": 10175 }, { "epoch": 1.294491794937031, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.083972930908203, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8717065453529358, "num_tokens": 388323971.0, "step": 10176 }, { "epoch": 1.2946190052156215, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.097870826721191, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8626014590263367, "num_tokens": 388362256.0, "step": 10177 }, { "epoch": 1.294746215494212, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.009132385253906, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8870134949684143, "num_tokens": 388401120.0, "step": 10178 }, { "epoch": 1.2948734257728025, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.071605682373047, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8693028688430786, "num_tokens": 388446833.0, "step": 10179 }, { "epoch": 1.2950006360513928, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.082799911499023, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.878846287727356, "num_tokens": 388482266.0, "step": 10180 }, { "epoch": 1.2951278463299833, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.116342544555664, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8683748245239258, "num_tokens": 388518024.0, "step": 10181 }, { "epoch": 1.2952550566085739, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.120887756347656, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8692230582237244, "num_tokens": 388564246.0, "step": 10182 }, { "epoch": 1.2953822668871644, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.081378936767578, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8707885146141052, "num_tokens": 388603556.0, "step": 10183 }, { "epoch": 1.295509477165755, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.141382217407227, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8751376867294312, "num_tokens": 388643615.0, "step": 10184 }, { "epoch": 1.2956366874443455, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.073476791381836, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.870348334312439, "num_tokens": 388682917.0, "step": 10185 }, { "epoch": 1.295763897722936, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.085356712341309, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8697634935379028, "num_tokens": 388720269.0, "step": 10186 }, { "epoch": 1.2958911080015265, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.121478080749512, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8595739006996155, "num_tokens": 388759245.0, "step": 10187 }, { "epoch": 1.296018318280117, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.070375442504883, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8635629415512085, "num_tokens": 388795673.0, "step": 10188 }, { "epoch": 1.2961455285587076, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 12.08354663848877, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8708000779151917, "num_tokens": 388832926.0, "step": 10189 }, { "epoch": 1.296272738837298, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.055425643920898, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8710973858833313, "num_tokens": 388869767.0, "step": 10190 }, { "epoch": 1.2963999491158886, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.085762023925781, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8694188594818115, "num_tokens": 388913347.0, "step": 10191 }, { "epoch": 1.2965271593944792, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.056546211242676, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8801079392433167, "num_tokens": 388948211.0, "step": 10192 }, { "epoch": 1.2966543696730697, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.0388822555542, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8721084594726562, "num_tokens": 388989198.0, "step": 10193 }, { "epoch": 1.2967815799516602, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.045938491821289, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.858946681022644, "num_tokens": 389029076.0, "step": 10194 }, { "epoch": 1.2969087902302507, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.116695404052734, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8675150871276855, "num_tokens": 389066751.0, "step": 10195 }, { "epoch": 1.297036000508841, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.051788330078125, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8652702569961548, "num_tokens": 389104207.0, "step": 10196 }, { "epoch": 1.2971632107874316, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.12227725982666, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8708886504173279, "num_tokens": 389141941.0, "step": 10197 }, { "epoch": 1.297290421066022, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.025239944458008, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8778566122055054, "num_tokens": 389185326.0, "step": 10198 }, { "epoch": 1.2974176313446126, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.132030487060547, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8772886991500854, "num_tokens": 389221512.0, "step": 10199 }, { "epoch": 1.2975448416232032, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.006586074829102, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8752787113189697, "num_tokens": 389259388.0, "step": 10200 }, { "epoch": 1.2976720519017937, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.066163063049316, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8566683530807495, "num_tokens": 389298773.0, "step": 10201 }, { "epoch": 1.2977992621803842, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.065686225891113, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8692842125892639, "num_tokens": 389339680.0, "step": 10202 }, { "epoch": 1.2979264724589747, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.102304458618164, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8707784414291382, "num_tokens": 389380755.0, "step": 10203 }, { "epoch": 1.2980536827375653, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.091598510742188, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8734371066093445, "num_tokens": 389412031.0, "step": 10204 }, { "epoch": 1.2981808930161556, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.122790336608887, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8773130774497986, "num_tokens": 389448299.0, "step": 10205 }, { "epoch": 1.298308103294746, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.166046142578125, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8623816967010498, "num_tokens": 389483962.0, "step": 10206 }, { "epoch": 1.2984353135733366, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.114205360412598, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8593474626541138, "num_tokens": 389521989.0, "step": 10207 }, { "epoch": 1.2985625238519272, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.096074104309082, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8625825643539429, "num_tokens": 389558228.0, "step": 10208 }, { "epoch": 1.2986897341305177, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.151333808898926, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8677152395248413, "num_tokens": 389597164.0, "step": 10209 }, { "epoch": 1.2988169444091082, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.08010482788086, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8687357902526855, "num_tokens": 389636435.0, "step": 10210 }, { "epoch": 1.2989441546876987, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.168035507202148, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8599314093589783, "num_tokens": 389670498.0, "step": 10211 }, { "epoch": 1.2990713649662893, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.108156204223633, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8700590133666992, "num_tokens": 389707433.0, "step": 10212 }, { "epoch": 1.2991985752448798, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.092182159423828, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8680503368377686, "num_tokens": 389747300.0, "step": 10213 }, { "epoch": 1.2993257855234703, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.066479682922363, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8621120452880859, "num_tokens": 389788381.0, "step": 10214 }, { "epoch": 1.2994529958020609, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.139497756958008, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.871275782585144, "num_tokens": 389821792.0, "step": 10215 }, { "epoch": 1.2995802060806514, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.12167739868164, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8794213533401489, "num_tokens": 389859370.0, "step": 10216 }, { "epoch": 1.299707416359242, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.051628112792969, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8595138788223267, "num_tokens": 389897020.0, "step": 10217 }, { "epoch": 1.2998346266378324, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.148253440856934, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8416266441345215, "num_tokens": 389927107.0, "step": 10218 }, { "epoch": 1.299961836916423, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.115498542785645, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8649126887321472, "num_tokens": 389966323.0, "step": 10219 }, { "epoch": 1.3000890471950133, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.107576370239258, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.867177426815033, "num_tokens": 390001416.0, "step": 10220 }, { "epoch": 1.3002162574736038, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.0835542678833, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8705214858055115, "num_tokens": 390042326.0, "step": 10221 }, { "epoch": 1.3003434677521943, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.054325103759766, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8586363792419434, "num_tokens": 390084443.0, "step": 10222 }, { "epoch": 1.3004706780307849, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.10350513458252, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8608033657073975, "num_tokens": 390126309.0, "step": 10223 }, { "epoch": 1.3005978883093754, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.068877220153809, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8723799586296082, "num_tokens": 390161463.0, "step": 10224 }, { "epoch": 1.300725098587966, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.060994148254395, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8782405257225037, "num_tokens": 390203221.0, "step": 10225 }, { "epoch": 1.3008523088665565, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.11382007598877, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8837032914161682, "num_tokens": 390243269.0, "step": 10226 }, { "epoch": 1.300979519145147, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.13889217376709, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.869371771812439, "num_tokens": 390282488.0, "step": 10227 }, { "epoch": 1.3011067294237375, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.168632507324219, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8630027174949646, "num_tokens": 390320268.0, "step": 10228 }, { "epoch": 1.3012339397023278, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.091639518737793, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8584111332893372, "num_tokens": 390354733.0, "step": 10229 }, { "epoch": 1.3013611499809183, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.088990211486816, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8673745393753052, "num_tokens": 390388161.0, "step": 10230 }, { "epoch": 1.3014883602595089, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.107576370239258, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8779125213623047, "num_tokens": 390424609.0, "step": 10231 }, { "epoch": 1.3016155705380994, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.118204116821289, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8841803669929504, "num_tokens": 390460809.0, "step": 10232 }, { "epoch": 1.30174278081669, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.063192367553711, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8572989702224731, "num_tokens": 390500198.0, "step": 10233 }, { "epoch": 1.3018699910952805, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.080916404724121, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8619195818901062, "num_tokens": 390545588.0, "step": 10234 }, { "epoch": 1.301997201373871, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.103875160217285, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8584247827529907, "num_tokens": 390581681.0, "step": 10235 }, { "epoch": 1.3021244116524615, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.091635704040527, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.884680449962616, "num_tokens": 390619122.0, "step": 10236 }, { "epoch": 1.302251621931052, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.078347206115723, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8557320237159729, "num_tokens": 390667974.0, "step": 10237 }, { "epoch": 1.3023788322096426, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.143030166625977, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8695001602172852, "num_tokens": 390705188.0, "step": 10238 }, { "epoch": 1.302506042488233, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.032600402832031, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8769348859786987, "num_tokens": 390746892.0, "step": 10239 }, { "epoch": 1.3026332527668236, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.052680015563965, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8823464512825012, "num_tokens": 390787727.0, "step": 10240 }, { "epoch": 1.3027604630454142, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.164327621459961, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8599570989608765, "num_tokens": 390823702.0, "step": 10241 }, { "epoch": 1.3028876733240047, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.04336929321289, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8668032288551331, "num_tokens": 390860256.0, "step": 10242 }, { "epoch": 1.3030148836025952, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.056135177612305, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8816776275634766, "num_tokens": 390901676.0, "step": 10243 }, { "epoch": 1.3031420938811857, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.067219734191895, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8511576652526855, "num_tokens": 390936911.0, "step": 10244 }, { "epoch": 1.303269304159776, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.10049819946289, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8559409379959106, "num_tokens": 390976185.0, "step": 10245 }, { "epoch": 1.3033965144383666, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.114411354064941, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8673602342605591, "num_tokens": 391014661.0, "step": 10246 }, { "epoch": 1.303523724716957, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.065702438354492, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8597131967544556, "num_tokens": 391056205.0, "step": 10247 }, { "epoch": 1.3036509349955476, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.119088172912598, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8693581223487854, "num_tokens": 391092187.0, "step": 10248 }, { "epoch": 1.3037781452741382, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.058238983154297, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8834616541862488, "num_tokens": 391130890.0, "step": 10249 }, { "epoch": 1.3039053555527287, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.117620468139648, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8775680661201477, "num_tokens": 391172783.0, "step": 10250 }, { "epoch": 1.3040325658313192, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.153450965881348, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8515521287918091, "num_tokens": 391215031.0, "step": 10251 }, { "epoch": 1.3041597761099097, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.108536720275879, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8707551956176758, "num_tokens": 391251400.0, "step": 10252 }, { "epoch": 1.3042869863885003, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.089037895202637, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.859021782875061, "num_tokens": 391285895.0, "step": 10253 }, { "epoch": 1.3044141966670906, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.116317749023438, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8694310188293457, "num_tokens": 391328361.0, "step": 10254 }, { "epoch": 1.304541406945681, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.028861999511719, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8666839599609375, "num_tokens": 391365019.0, "step": 10255 }, { "epoch": 1.3046686172242716, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.129892349243164, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8521628379821777, "num_tokens": 391400407.0, "step": 10256 }, { "epoch": 1.3047958275028622, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.134673118591309, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8502324819564819, "num_tokens": 391440660.0, "step": 10257 }, { "epoch": 1.3049230377814527, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.063421249389648, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8694251775741577, "num_tokens": 391481096.0, "step": 10258 }, { "epoch": 1.3050502480600432, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.126962661743164, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8680645823478699, "num_tokens": 391516356.0, "step": 10259 }, { "epoch": 1.3051774583386337, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.090703964233398, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8560811281204224, "num_tokens": 391554691.0, "step": 10260 }, { "epoch": 1.3053046686172243, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.084305763244629, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8538931012153625, "num_tokens": 391592473.0, "step": 10261 }, { "epoch": 1.3054318788958148, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.098630905151367, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.872997522354126, "num_tokens": 391632091.0, "step": 10262 }, { "epoch": 1.3055590891744053, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.059104919433594, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8602094650268555, "num_tokens": 391676121.0, "step": 10263 }, { "epoch": 1.3056862994529959, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.012409210205078, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8655651807785034, "num_tokens": 391709664.0, "step": 10264 }, { "epoch": 1.3058135097315864, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.139779090881348, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8558763265609741, "num_tokens": 391755828.0, "step": 10265 }, { "epoch": 1.305940720010177, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.029050827026367, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8835102319717407, "num_tokens": 391800232.0, "step": 10266 }, { "epoch": 1.3060679302887674, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.05379867553711, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8687968850135803, "num_tokens": 391837612.0, "step": 10267 }, { "epoch": 1.306195140567358, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.078862190246582, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8775040507316589, "num_tokens": 391871193.0, "step": 10268 }, { "epoch": 1.3063223508459483, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.17876148223877, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8590414524078369, "num_tokens": 391908203.0, "step": 10269 }, { "epoch": 1.3064495611245388, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.055807113647461, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8729572296142578, "num_tokens": 391938744.0, "step": 10270 }, { "epoch": 1.3065767714031293, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.1193208694458, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8696722388267517, "num_tokens": 391976020.0, "step": 10271 }, { "epoch": 1.3067039816817199, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.091991424560547, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8585588932037354, "num_tokens": 392018006.0, "step": 10272 }, { "epoch": 1.3068311919603104, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.070411682128906, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8653653860092163, "num_tokens": 392055358.0, "step": 10273 }, { "epoch": 1.306958402238901, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.071484565734863, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8656297922134399, "num_tokens": 392097727.0, "step": 10274 }, { "epoch": 1.3070856125174914, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.069445610046387, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8728755712509155, "num_tokens": 392144641.0, "step": 10275 }, { "epoch": 1.307212822796082, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.094863891601562, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.875086784362793, "num_tokens": 392178810.0, "step": 10276 }, { "epoch": 1.3073400330746725, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.17564582824707, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8621264696121216, "num_tokens": 392213571.0, "step": 10277 }, { "epoch": 1.3074672433532628, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.146620750427246, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8631042838096619, "num_tokens": 392254628.0, "step": 10278 }, { "epoch": 1.3075944536318533, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.118998527526855, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8660475015640259, "num_tokens": 392297243.0, "step": 10279 }, { "epoch": 1.3077216639104439, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.119515419006348, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8690608739852905, "num_tokens": 392329116.0, "step": 10280 }, { "epoch": 1.3078488741890344, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.170600891113281, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8642169237136841, "num_tokens": 392370542.0, "step": 10281 }, { "epoch": 1.307976084467625, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.093832015991211, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8719963431358337, "num_tokens": 392411155.0, "step": 10282 }, { "epoch": 1.3081032947462155, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.151998519897461, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8675310611724854, "num_tokens": 392450440.0, "step": 10283 }, { "epoch": 1.308230505024806, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.114784240722656, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8773364424705505, "num_tokens": 392488545.0, "step": 10284 }, { "epoch": 1.3083577153033965, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.127521514892578, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8667561411857605, "num_tokens": 392527138.0, "step": 10285 }, { "epoch": 1.308484925581987, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.133594512939453, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8674178123474121, "num_tokens": 392568181.0, "step": 10286 }, { "epoch": 1.3086121358605776, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.127872467041016, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.882175624370575, "num_tokens": 392610900.0, "step": 10287 }, { "epoch": 1.308739346139168, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.098190307617188, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8590364456176758, "num_tokens": 392646624.0, "step": 10288 }, { "epoch": 1.3088665564177586, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.180060386657715, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8678112626075745, "num_tokens": 392683103.0, "step": 10289 }, { "epoch": 1.3089937666963491, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.123859405517578, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8685516119003296, "num_tokens": 392723169.0, "step": 10290 }, { "epoch": 1.3091209769749397, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.138473510742188, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8822460174560547, "num_tokens": 392762323.0, "step": 10291 }, { "epoch": 1.3092481872535302, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.166879653930664, "learning_rate": 1e-06, "loss": 0.3494, "mean_token_accuracy": 0.8840745687484741, "num_tokens": 392797555.0, "step": 10292 }, { "epoch": 1.3093753975321207, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.053311347961426, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8780637383460999, "num_tokens": 392840001.0, "step": 10293 }, { "epoch": 1.309502607810711, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.146687507629395, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8667755126953125, "num_tokens": 392880871.0, "step": 10294 }, { "epoch": 1.3096298180893016, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.137469291687012, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8784324526786804, "num_tokens": 392917554.0, "step": 10295 }, { "epoch": 1.309757028367892, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.111207962036133, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8848167657852173, "num_tokens": 392958601.0, "step": 10296 }, { "epoch": 1.3098842386464826, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.120573997497559, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8696876168251038, "num_tokens": 393000108.0, "step": 10297 }, { "epoch": 1.3100114489250732, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.135015487670898, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.8884527683258057, "num_tokens": 393032321.0, "step": 10298 }, { "epoch": 1.3101386592036637, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.163222312927246, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8663430213928223, "num_tokens": 393068347.0, "step": 10299 }, { "epoch": 1.3102658694822542, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.118003845214844, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8853756785392761, "num_tokens": 393102509.0, "step": 10300 }, { "epoch": 1.3103930797608447, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.115906715393066, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8766398429870605, "num_tokens": 393140974.0, "step": 10301 }, { "epoch": 1.3105202900394353, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.141650199890137, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8768340349197388, "num_tokens": 393177214.0, "step": 10302 }, { "epoch": 1.3106475003180256, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.146672248840332, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.87164306640625, "num_tokens": 393215801.0, "step": 10303 }, { "epoch": 1.310774710596616, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.060752868652344, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.879368782043457, "num_tokens": 393256444.0, "step": 10304 }, { "epoch": 1.3109019208752066, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.191871643066406, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8645139336585999, "num_tokens": 393300238.0, "step": 10305 }, { "epoch": 1.3110291311537972, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.046138763427734, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8548848032951355, "num_tokens": 393344087.0, "step": 10306 }, { "epoch": 1.3111563414323877, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.20084285736084, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8728092908859253, "num_tokens": 393381678.0, "step": 10307 }, { "epoch": 1.3112835517109782, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.095352172851562, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8666999340057373, "num_tokens": 393415740.0, "step": 10308 }, { "epoch": 1.3114107619895687, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.140910148620605, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8444757461547852, "num_tokens": 393457692.0, "step": 10309 }, { "epoch": 1.3115379722681593, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.129958152770996, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8637853860855103, "num_tokens": 393499003.0, "step": 10310 }, { "epoch": 1.3116651825467498, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.142719268798828, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8577871918678284, "num_tokens": 393533842.0, "step": 10311 }, { "epoch": 1.3117923928253403, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.146190643310547, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.875594973564148, "num_tokens": 393573906.0, "step": 10312 }, { "epoch": 1.3119196031039309, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.178961753845215, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8663832545280457, "num_tokens": 393615298.0, "step": 10313 }, { "epoch": 1.3120468133825214, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.126691818237305, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8579983711242676, "num_tokens": 393654690.0, "step": 10314 }, { "epoch": 1.312174023661112, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.130393028259277, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8706393241882324, "num_tokens": 393696222.0, "step": 10315 }, { "epoch": 1.3123012339397024, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.143038749694824, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8644641637802124, "num_tokens": 393736748.0, "step": 10316 }, { "epoch": 1.312428444218293, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.136188507080078, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8593074083328247, "num_tokens": 393773077.0, "step": 10317 }, { "epoch": 1.3125556544968833, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.153742790222168, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8796016573905945, "num_tokens": 393806466.0, "step": 10318 }, { "epoch": 1.3126828647754738, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.130322456359863, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8644338250160217, "num_tokens": 393840922.0, "step": 10319 }, { "epoch": 1.3128100750540643, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.142996788024902, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8730406165122986, "num_tokens": 393874544.0, "step": 10320 }, { "epoch": 1.3129372853326549, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.143990516662598, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8649327754974365, "num_tokens": 393917773.0, "step": 10321 }, { "epoch": 1.3130644956112454, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.173135757446289, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8638414144515991, "num_tokens": 393951525.0, "step": 10322 }, { "epoch": 1.313191705889836, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.110821723937988, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8728739023208618, "num_tokens": 393989796.0, "step": 10323 }, { "epoch": 1.3133189161684264, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.152414321899414, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8656185865402222, "num_tokens": 394028746.0, "step": 10324 }, { "epoch": 1.313446126447017, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.135589599609375, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8767490386962891, "num_tokens": 394065241.0, "step": 10325 }, { "epoch": 1.3135733367256075, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.15674114227295, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8626686334609985, "num_tokens": 394110113.0, "step": 10326 }, { "epoch": 1.3137005470041978, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.159205436706543, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8512202501296997, "num_tokens": 394146577.0, "step": 10327 }, { "epoch": 1.3138277572827883, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.16020393371582, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8679437637329102, "num_tokens": 394185392.0, "step": 10328 }, { "epoch": 1.3139549675613789, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.102418899536133, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8587889671325684, "num_tokens": 394226301.0, "step": 10329 }, { "epoch": 1.3140821778399694, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.196600914001465, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.858683705329895, "num_tokens": 394262654.0, "step": 10330 }, { "epoch": 1.31420938811856, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.145076751708984, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8801953196525574, "num_tokens": 394299185.0, "step": 10331 }, { "epoch": 1.3143365983971504, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.137524604797363, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8692694902420044, "num_tokens": 394331364.0, "step": 10332 }, { "epoch": 1.314463808675741, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.130993843078613, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8559019565582275, "num_tokens": 394375479.0, "step": 10333 }, { "epoch": 1.3145910189543315, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.209456443786621, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8734604716300964, "num_tokens": 394414913.0, "step": 10334 }, { "epoch": 1.314718229232922, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.158012390136719, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8693552017211914, "num_tokens": 394450889.0, "step": 10335 }, { "epoch": 1.3148454395115126, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.147967338562012, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8511910438537598, "num_tokens": 394495280.0, "step": 10336 }, { "epoch": 1.314972649790103, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.126579284667969, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.876303493976593, "num_tokens": 394536811.0, "step": 10337 }, { "epoch": 1.3150998600686936, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.156834602355957, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8656867742538452, "num_tokens": 394573353.0, "step": 10338 }, { "epoch": 1.3152270703472841, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.175766944885254, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.861061155796051, "num_tokens": 394603517.0, "step": 10339 }, { "epoch": 1.3153542806258747, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.1315279006958, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8739674091339111, "num_tokens": 394650547.0, "step": 10340 }, { "epoch": 1.3154814909044652, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.122535705566406, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8717052936553955, "num_tokens": 394684460.0, "step": 10341 }, { "epoch": 1.3156087011830557, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.153337478637695, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8830571174621582, "num_tokens": 394733520.0, "step": 10342 }, { "epoch": 1.315735911461646, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.092187881469727, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8620890974998474, "num_tokens": 394775498.0, "step": 10343 }, { "epoch": 1.3158631217402366, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.237149238586426, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8550845384597778, "num_tokens": 394815138.0, "step": 10344 }, { "epoch": 1.315990332018827, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.115216255187988, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8591443300247192, "num_tokens": 394850447.0, "step": 10345 }, { "epoch": 1.3161175422974176, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.162941932678223, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8612959980964661, "num_tokens": 394891800.0, "step": 10346 }, { "epoch": 1.3162447525760081, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.194985389709473, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8748186826705933, "num_tokens": 394930087.0, "step": 10347 }, { "epoch": 1.3163719628545987, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.11747932434082, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8770543932914734, "num_tokens": 394964497.0, "step": 10348 }, { "epoch": 1.3164991731331892, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.098458290100098, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8500291109085083, "num_tokens": 395003127.0, "step": 10349 }, { "epoch": 1.3166263834117797, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.150466918945312, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8568387031555176, "num_tokens": 395040570.0, "step": 10350 }, { "epoch": 1.3167535936903703, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.119564056396484, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8678227663040161, "num_tokens": 395075205.0, "step": 10351 }, { "epoch": 1.3168808039689606, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.148469924926758, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8677326440811157, "num_tokens": 395117492.0, "step": 10352 }, { "epoch": 1.317008014247551, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.138574600219727, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8647503852844238, "num_tokens": 395162449.0, "step": 10353 }, { "epoch": 1.3171352245261416, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.136116981506348, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8765109181404114, "num_tokens": 395197310.0, "step": 10354 }, { "epoch": 1.3172624348047322, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.09024429321289, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8788295984268188, "num_tokens": 395232420.0, "step": 10355 }, { "epoch": 1.3173896450833227, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.169221878051758, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8470363616943359, "num_tokens": 395269392.0, "step": 10356 }, { "epoch": 1.3175168553619132, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.136821746826172, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8562401533126831, "num_tokens": 395305496.0, "step": 10357 }, { "epoch": 1.3176440656405037, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.120336532592773, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8761768341064453, "num_tokens": 395347089.0, "step": 10358 }, { "epoch": 1.3177712759190943, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.082676887512207, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8777966499328613, "num_tokens": 395382885.0, "step": 10359 }, { "epoch": 1.3178984861976848, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.158896446228027, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8813211917877197, "num_tokens": 395416260.0, "step": 10360 }, { "epoch": 1.3180256964762753, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.09340763092041, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8585772514343262, "num_tokens": 395454179.0, "step": 10361 }, { "epoch": 1.3181529067548658, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.182005882263184, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8580480813980103, "num_tokens": 395489173.0, "step": 10362 }, { "epoch": 1.3182801170334564, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.142498970031738, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8617367148399353, "num_tokens": 395532057.0, "step": 10363 }, { "epoch": 1.318407327312047, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.18865966796875, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8732455968856812, "num_tokens": 395570321.0, "step": 10364 }, { "epoch": 1.3185345375906374, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.113988876342773, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8870958685874939, "num_tokens": 395609041.0, "step": 10365 }, { "epoch": 1.318661747869228, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.159942626953125, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8448182344436646, "num_tokens": 395649961.0, "step": 10366 }, { "epoch": 1.3187889581478183, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.098854064941406, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8688362240791321, "num_tokens": 395691882.0, "step": 10367 }, { "epoch": 1.3189161684264088, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.14436149597168, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8556411862373352, "num_tokens": 395731496.0, "step": 10368 }, { "epoch": 1.3190433787049993, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.110807418823242, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8717401027679443, "num_tokens": 395772618.0, "step": 10369 }, { "epoch": 1.3191705889835899, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.065741539001465, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8511911630630493, "num_tokens": 395811907.0, "step": 10370 }, { "epoch": 1.3192977992621804, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.191484451293945, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8634024858474731, "num_tokens": 395851186.0, "step": 10371 }, { "epoch": 1.319425009540771, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.054887771606445, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8678569793701172, "num_tokens": 395893868.0, "step": 10372 }, { "epoch": 1.3195522198193614, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.142664909362793, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8774324655532837, "num_tokens": 395930519.0, "step": 10373 }, { "epoch": 1.319679430097952, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.153116226196289, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8834901452064514, "num_tokens": 395965692.0, "step": 10374 }, { "epoch": 1.3198066403765425, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.197973251342773, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8685970902442932, "num_tokens": 396002316.0, "step": 10375 }, { "epoch": 1.3199338506551328, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.09068775177002, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.874724805355072, "num_tokens": 396043392.0, "step": 10376 }, { "epoch": 1.3200610609337233, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.155200004577637, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8645555973052979, "num_tokens": 396078603.0, "step": 10377 }, { "epoch": 1.3201882712123139, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.142919540405273, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.858506441116333, "num_tokens": 396121132.0, "step": 10378 }, { "epoch": 1.3203154814909044, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.190153121948242, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8768370151519775, "num_tokens": 396155730.0, "step": 10379 }, { "epoch": 1.320442691769495, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.17137336730957, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8532613515853882, "num_tokens": 396190976.0, "step": 10380 }, { "epoch": 1.3205699020480854, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.179784774780273, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8663673400878906, "num_tokens": 396229571.0, "step": 10381 }, { "epoch": 1.320697112326676, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.161964416503906, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8666502833366394, "num_tokens": 396265586.0, "step": 10382 }, { "epoch": 1.3208243226052665, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.177703857421875, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8722391128540039, "num_tokens": 396303810.0, "step": 10383 }, { "epoch": 1.320951532883857, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.133975982666016, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8870872259140015, "num_tokens": 396345671.0, "step": 10384 }, { "epoch": 1.3210787431624476, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.135123252868652, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8840224146842957, "num_tokens": 396388062.0, "step": 10385 }, { "epoch": 1.321205953441038, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.091282844543457, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8811521530151367, "num_tokens": 396425189.0, "step": 10386 }, { "epoch": 1.3213331637196286, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.09053897857666, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8582314252853394, "num_tokens": 396461861.0, "step": 10387 }, { "epoch": 1.3214603739982191, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.143287658691406, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8788527250289917, "num_tokens": 396506190.0, "step": 10388 }, { "epoch": 1.3215875842768097, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.199857711791992, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8693019151687622, "num_tokens": 396538900.0, "step": 10389 }, { "epoch": 1.3217147945554002, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.069199562072754, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8773965239524841, "num_tokens": 396577029.0, "step": 10390 }, { "epoch": 1.3218420048339907, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.092846870422363, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8530767560005188, "num_tokens": 396616917.0, "step": 10391 }, { "epoch": 1.321969215112581, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.211223602294922, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8772293925285339, "num_tokens": 396651726.0, "step": 10392 }, { "epoch": 1.3220964253911716, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.156237602233887, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8560850620269775, "num_tokens": 396693796.0, "step": 10393 }, { "epoch": 1.322223635669762, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.149720191955566, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8579210638999939, "num_tokens": 396731947.0, "step": 10394 }, { "epoch": 1.3223508459483526, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.215818405151367, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.870669960975647, "num_tokens": 396767062.0, "step": 10395 }, { "epoch": 1.3224780562269431, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.121769905090332, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8720956444740295, "num_tokens": 396803918.0, "step": 10396 }, { "epoch": 1.3226052665055337, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.162466049194336, "learning_rate": 1e-06, "loss": 0.4999, "mean_token_accuracy": 0.8411715030670166, "num_tokens": 396850314.0, "step": 10397 }, { "epoch": 1.3227324767841242, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.169349670410156, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8611716032028198, "num_tokens": 396887920.0, "step": 10398 }, { "epoch": 1.3228596870627147, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.167326927185059, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8574909567832947, "num_tokens": 396923369.0, "step": 10399 }, { "epoch": 1.3229868973413053, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.135092735290527, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8661705255508423, "num_tokens": 396962338.0, "step": 10400 }, { "epoch": 1.3231141076198956, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.133125305175781, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8784151673316956, "num_tokens": 396998275.0, "step": 10401 }, { "epoch": 1.323241317898486, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.137685775756836, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8655716180801392, "num_tokens": 397040034.0, "step": 10402 }, { "epoch": 1.3233685281770766, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.101944923400879, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8863211870193481, "num_tokens": 397080549.0, "step": 10403 }, { "epoch": 1.3234957384556671, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.173027038574219, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8802055716514587, "num_tokens": 397120803.0, "step": 10404 }, { "epoch": 1.3236229487342577, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.110067367553711, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8716514110565186, "num_tokens": 397160463.0, "step": 10405 }, { "epoch": 1.3237501590128482, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.145830154418945, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8611952066421509, "num_tokens": 397205425.0, "step": 10406 }, { "epoch": 1.3238773692914387, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.14124870300293, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8636254072189331, "num_tokens": 397238907.0, "step": 10407 }, { "epoch": 1.3240045795700293, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.157249450683594, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8673831820487976, "num_tokens": 397281639.0, "step": 10408 }, { "epoch": 1.3241317898486198, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.131535530090332, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8566798567771912, "num_tokens": 397323350.0, "step": 10409 }, { "epoch": 1.3242590001272103, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.164003372192383, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.861299991607666, "num_tokens": 397364438.0, "step": 10410 }, { "epoch": 1.3243862104058008, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.15899658203125, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8553599119186401, "num_tokens": 397406881.0, "step": 10411 }, { "epoch": 1.3245134206843914, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.117844581604004, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.868363082408905, "num_tokens": 397451190.0, "step": 10412 }, { "epoch": 1.324640630962982, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.175409317016602, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8695828318595886, "num_tokens": 397491063.0, "step": 10413 }, { "epoch": 1.3247678412415724, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.153939247131348, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8688356876373291, "num_tokens": 397529961.0, "step": 10414 }, { "epoch": 1.324895051520163, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.171259880065918, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8784071207046509, "num_tokens": 397565616.0, "step": 10415 }, { "epoch": 1.3250222617987533, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.107199668884277, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8744056820869446, "num_tokens": 397602223.0, "step": 10416 }, { "epoch": 1.3251494720773438, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.163900375366211, "learning_rate": 1e-06, "loss": 0.4807, "mean_token_accuracy": 0.84660404920578, "num_tokens": 397638910.0, "step": 10417 }, { "epoch": 1.3252766823559343, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.125178337097168, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8603025674819946, "num_tokens": 397681200.0, "step": 10418 }, { "epoch": 1.3254038926345248, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.092503547668457, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8703938722610474, "num_tokens": 397724832.0, "step": 10419 }, { "epoch": 1.3255311029131154, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.134183883666992, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8606990575790405, "num_tokens": 397765537.0, "step": 10420 }, { "epoch": 1.325658313191706, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.098603248596191, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8588346242904663, "num_tokens": 397800945.0, "step": 10421 }, { "epoch": 1.3257855234702964, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.136902809143066, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8741147518157959, "num_tokens": 397840786.0, "step": 10422 }, { "epoch": 1.325912733748887, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.062459945678711, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8724691271781921, "num_tokens": 397875607.0, "step": 10423 }, { "epoch": 1.3260399440274775, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.098219871520996, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8840705752372742, "num_tokens": 397909524.0, "step": 10424 }, { "epoch": 1.3261671543060678, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.125308990478516, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8524078726768494, "num_tokens": 397947862.0, "step": 10425 }, { "epoch": 1.3262943645846583, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.136955261230469, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8814408779144287, "num_tokens": 397980958.0, "step": 10426 }, { "epoch": 1.3264215748632489, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.11678409576416, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8653891086578369, "num_tokens": 398021231.0, "step": 10427 }, { "epoch": 1.3265487851418394, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.119407653808594, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8657580018043518, "num_tokens": 398059202.0, "step": 10428 }, { "epoch": 1.32667599542043, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.142230033874512, "learning_rate": 1e-06, "loss": 0.3407, "mean_token_accuracy": 0.889612078666687, "num_tokens": 398097121.0, "step": 10429 }, { "epoch": 1.3268032056990204, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.153607368469238, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8671760559082031, "num_tokens": 398130413.0, "step": 10430 }, { "epoch": 1.326930415977611, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.085704803466797, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8664661645889282, "num_tokens": 398172535.0, "step": 10431 }, { "epoch": 1.3270576262562015, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.201457977294922, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8699911236763, "num_tokens": 398206681.0, "step": 10432 }, { "epoch": 1.327184836534792, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.120166778564453, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8714697360992432, "num_tokens": 398245206.0, "step": 10433 }, { "epoch": 1.3273120468133826, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168774604797363, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8649307489395142, "num_tokens": 398278902.0, "step": 10434 }, { "epoch": 1.327439257091973, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.175368309020996, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8685306310653687, "num_tokens": 398315833.0, "step": 10435 }, { "epoch": 1.3275664673705636, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.14554214477539, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8507269620895386, "num_tokens": 398356496.0, "step": 10436 }, { "epoch": 1.3276936776491541, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.16930103302002, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8735519647598267, "num_tokens": 398393661.0, "step": 10437 }, { "epoch": 1.3278208879277447, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.193456649780273, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8847546577453613, "num_tokens": 398430463.0, "step": 10438 }, { "epoch": 1.3279480982063352, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.141948699951172, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8668341636657715, "num_tokens": 398470382.0, "step": 10439 }, { "epoch": 1.3280753084849257, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.192182540893555, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8513234257698059, "num_tokens": 398504793.0, "step": 10440 }, { "epoch": 1.328202518763516, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168495178222656, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8591022491455078, "num_tokens": 398541038.0, "step": 10441 }, { "epoch": 1.3283297290421066, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.225229263305664, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8510462641716003, "num_tokens": 398578047.0, "step": 10442 }, { "epoch": 1.328456939320697, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.163456916809082, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8572427034378052, "num_tokens": 398618946.0, "step": 10443 }, { "epoch": 1.3285841495992876, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.10977554321289, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8720600008964539, "num_tokens": 398656895.0, "step": 10444 }, { "epoch": 1.3287113598778781, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.208579063415527, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8676227927207947, "num_tokens": 398696521.0, "step": 10445 }, { "epoch": 1.3288385701564687, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.154114723205566, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8685002326965332, "num_tokens": 398735275.0, "step": 10446 }, { "epoch": 1.3289657804350592, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.12792682647705, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8533518314361572, "num_tokens": 398774753.0, "step": 10447 }, { "epoch": 1.3290929907136497, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.197613716125488, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8621206879615784, "num_tokens": 398810893.0, "step": 10448 }, { "epoch": 1.3292202009922403, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.094369888305664, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8623241782188416, "num_tokens": 398849741.0, "step": 10449 }, { "epoch": 1.3293474112708306, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.20010757446289, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8618054986000061, "num_tokens": 398881186.0, "step": 10450 }, { "epoch": 1.329474621549421, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.11996841430664, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8847231864929199, "num_tokens": 398918677.0, "step": 10451 }, { "epoch": 1.3296018318280116, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.124969482421875, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8648197054862976, "num_tokens": 398953435.0, "step": 10452 }, { "epoch": 1.3297290421066021, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.159732818603516, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8747868537902832, "num_tokens": 398991799.0, "step": 10453 }, { "epoch": 1.3298562523851927, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.135793685913086, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.871637761592865, "num_tokens": 399023211.0, "step": 10454 }, { "epoch": 1.3299834626637832, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.206326484680176, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8641321659088135, "num_tokens": 399061431.0, "step": 10455 }, { "epoch": 1.3301106729423737, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.121399879455566, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8695144653320312, "num_tokens": 399097384.0, "step": 10456 }, { "epoch": 1.3302378832209643, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.111822128295898, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8705812096595764, "num_tokens": 399133935.0, "step": 10457 }, { "epoch": 1.3303650934995548, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.169880867004395, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8726025223731995, "num_tokens": 399165709.0, "step": 10458 }, { "epoch": 1.3304923037781453, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.1271390914917, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8655219674110413, "num_tokens": 399203548.0, "step": 10459 }, { "epoch": 1.3306195140567358, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.180115699768066, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8657199144363403, "num_tokens": 399244341.0, "step": 10460 }, { "epoch": 1.3307467243353264, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.173111915588379, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8724542260169983, "num_tokens": 399282284.0, "step": 10461 }, { "epoch": 1.330873934613917, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.121848106384277, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8809959292411804, "num_tokens": 399320523.0, "step": 10462 }, { "epoch": 1.3310011448925074, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.1935396194458, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8680053949356079, "num_tokens": 399357489.0, "step": 10463 }, { "epoch": 1.331128355171098, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.148780822753906, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8555042743682861, "num_tokens": 399393561.0, "step": 10464 }, { "epoch": 1.3312555654496883, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.137778282165527, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8735195994377136, "num_tokens": 399424514.0, "step": 10465 }, { "epoch": 1.3313827757282788, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.19144344329834, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8735377788543701, "num_tokens": 399461372.0, "step": 10466 }, { "epoch": 1.3315099860068693, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.116033554077148, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8818484544754028, "num_tokens": 399502503.0, "step": 10467 }, { "epoch": 1.3316371962854598, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.205060005187988, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.862262487411499, "num_tokens": 399537836.0, "step": 10468 }, { "epoch": 1.3317644065640504, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.141794204711914, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8642154335975647, "num_tokens": 399575189.0, "step": 10469 }, { "epoch": 1.331891616842641, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.144338607788086, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8494017720222473, "num_tokens": 399614406.0, "step": 10470 }, { "epoch": 1.3320188271212314, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.18139934539795, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8730165958404541, "num_tokens": 399652847.0, "step": 10471 }, { "epoch": 1.332146037399822, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.153153419494629, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8581625819206238, "num_tokens": 399691628.0, "step": 10472 }, { "epoch": 1.3322732476784125, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.15397834777832, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8780991435050964, "num_tokens": 399732875.0, "step": 10473 }, { "epoch": 1.3324004579570028, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.150967597961426, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8790676593780518, "num_tokens": 399769537.0, "step": 10474 }, { "epoch": 1.3325276682355933, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.14028549194336, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.851942777633667, "num_tokens": 399807757.0, "step": 10475 }, { "epoch": 1.3326548785141838, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.176566123962402, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8747027516365051, "num_tokens": 399837351.0, "step": 10476 }, { "epoch": 1.3327820887927744, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.124312400817871, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8862026929855347, "num_tokens": 399873678.0, "step": 10477 }, { "epoch": 1.332909299071365, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.143972396850586, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8700218200683594, "num_tokens": 399916206.0, "step": 10478 }, { "epoch": 1.3330365093499554, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.164653778076172, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8793231844902039, "num_tokens": 399954144.0, "step": 10479 }, { "epoch": 1.333163719628546, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.165283203125, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8757782578468323, "num_tokens": 399995865.0, "step": 10480 }, { "epoch": 1.3332909299071365, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.161592483520508, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8695162534713745, "num_tokens": 400033234.0, "step": 10481 }, { "epoch": 1.333418140185727, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.199504852294922, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.876671552658081, "num_tokens": 400066842.0, "step": 10482 }, { "epoch": 1.3335453504643175, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.131319999694824, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8795725107192993, "num_tokens": 400099380.0, "step": 10483 }, { "epoch": 1.333672560742908, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.160697937011719, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8805307149887085, "num_tokens": 400137419.0, "step": 10484 }, { "epoch": 1.3337997710214986, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.181686401367188, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8692684173583984, "num_tokens": 400171086.0, "step": 10485 }, { "epoch": 1.3339269813000891, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.169557571411133, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8661659359931946, "num_tokens": 400206763.0, "step": 10486 }, { "epoch": 1.3340541915786797, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.162680625915527, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8532426357269287, "num_tokens": 400242708.0, "step": 10487 }, { "epoch": 1.3341814018572702, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.181290626525879, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8483175039291382, "num_tokens": 400277417.0, "step": 10488 }, { "epoch": 1.3343086121358605, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.16337776184082, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.868362307548523, "num_tokens": 400315811.0, "step": 10489 }, { "epoch": 1.334435822414451, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.167970657348633, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.848540186882019, "num_tokens": 400356321.0, "step": 10490 }, { "epoch": 1.3345630326930416, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.170933723449707, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8683041334152222, "num_tokens": 400392382.0, "step": 10491 }, { "epoch": 1.334690242971632, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.135249137878418, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8752250671386719, "num_tokens": 400434947.0, "step": 10492 }, { "epoch": 1.3348174532502226, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.12120246887207, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8497426509857178, "num_tokens": 400472742.0, "step": 10493 }, { "epoch": 1.3349446635288131, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.181354522705078, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8819552063941956, "num_tokens": 400510029.0, "step": 10494 }, { "epoch": 1.3350718738074037, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168139457702637, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8656069040298462, "num_tokens": 400547383.0, "step": 10495 }, { "epoch": 1.3351990840859942, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.086932182312012, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8787120580673218, "num_tokens": 400585242.0, "step": 10496 }, { "epoch": 1.3353262943645847, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.203911781311035, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8792989253997803, "num_tokens": 400623117.0, "step": 10497 }, { "epoch": 1.3354535046431752, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.126178741455078, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8577847480773926, "num_tokens": 400660170.0, "step": 10498 }, { "epoch": 1.3355807149217656, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.143081665039062, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8741123676300049, "num_tokens": 400700467.0, "step": 10499 }, { "epoch": 1.335707925200356, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.141761779785156, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8628581166267395, "num_tokens": 400740269.0, "step": 10500 }, { "epoch": 1.3358351354789466, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.21703052520752, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8778198957443237, "num_tokens": 400776465.0, "step": 10501 }, { "epoch": 1.3359623457575371, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.100783348083496, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8663088083267212, "num_tokens": 400811210.0, "step": 10502 }, { "epoch": 1.3360895560361277, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.237525939941406, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8615567088127136, "num_tokens": 400848820.0, "step": 10503 }, { "epoch": 1.3362167663147182, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.178728103637695, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8693954348564148, "num_tokens": 400888774.0, "step": 10504 }, { "epoch": 1.3363439765933087, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.16004753112793, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8709299564361572, "num_tokens": 400935067.0, "step": 10505 }, { "epoch": 1.3364711868718993, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.110660552978516, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8780825734138489, "num_tokens": 400974122.0, "step": 10506 }, { "epoch": 1.3365983971504898, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.172593116760254, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8779648542404175, "num_tokens": 401008911.0, "step": 10507 }, { "epoch": 1.3367256074290803, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.224441528320312, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8744053840637207, "num_tokens": 401039453.0, "step": 10508 }, { "epoch": 1.3368528177076708, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.142518997192383, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8777824640274048, "num_tokens": 401075854.0, "step": 10509 }, { "epoch": 1.3369800279862614, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.190777778625488, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.866456151008606, "num_tokens": 401114922.0, "step": 10510 }, { "epoch": 1.337107238264852, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.181256294250488, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8772152066230774, "num_tokens": 401155833.0, "step": 10511 }, { "epoch": 1.3372344485434424, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.123738288879395, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8748896718025208, "num_tokens": 401196971.0, "step": 10512 }, { "epoch": 1.337361658822033, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.177312850952148, "learning_rate": 1e-06, "loss": 0.5092, "mean_token_accuracy": 0.8379693031311035, "num_tokens": 401237312.0, "step": 10513 }, { "epoch": 1.3374888691006233, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.15467357635498, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8658775687217712, "num_tokens": 401274630.0, "step": 10514 }, { "epoch": 1.3376160793792138, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.181120872497559, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8733745813369751, "num_tokens": 401311780.0, "step": 10515 }, { "epoch": 1.3377432896578043, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.190750122070312, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8762595653533936, "num_tokens": 401352614.0, "step": 10516 }, { "epoch": 1.3378704999363948, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.082911491394043, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8716368675231934, "num_tokens": 401394495.0, "step": 10517 }, { "epoch": 1.3379977102149854, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.1983642578125, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8550550937652588, "num_tokens": 401427758.0, "step": 10518 }, { "epoch": 1.338124920493576, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.11034870147705, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8539448976516724, "num_tokens": 401473868.0, "step": 10519 }, { "epoch": 1.3382521307721664, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.155200004577637, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8516860604286194, "num_tokens": 401510459.0, "step": 10520 }, { "epoch": 1.338379341050757, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.190651893615723, "learning_rate": 1e-06, "loss": 0.5072, "mean_token_accuracy": 0.8408747315406799, "num_tokens": 401545283.0, "step": 10521 }, { "epoch": 1.3385065513293475, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.152914047241211, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8700979351997375, "num_tokens": 401577166.0, "step": 10522 }, { "epoch": 1.3386337616079378, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.151849746704102, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8813114762306213, "num_tokens": 401619075.0, "step": 10523 }, { "epoch": 1.3387609718865283, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105512619018555, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8621066808700562, "num_tokens": 401661286.0, "step": 10524 }, { "epoch": 1.3388881821651188, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.197957038879395, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8592916131019592, "num_tokens": 401699672.0, "step": 10525 }, { "epoch": 1.3390153924437094, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.118165016174316, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8615542650222778, "num_tokens": 401732676.0, "step": 10526 }, { "epoch": 1.3391426027223, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.20991325378418, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8669330477714539, "num_tokens": 401771593.0, "step": 10527 }, { "epoch": 1.3392698130008904, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.10847282409668, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8549035787582397, "num_tokens": 401809605.0, "step": 10528 }, { "epoch": 1.339397023279481, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.176745414733887, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.858447790145874, "num_tokens": 401845349.0, "step": 10529 }, { "epoch": 1.3395242335580715, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.13963794708252, "learning_rate": 1e-06, "loss": 0.3311, "mean_token_accuracy": 0.8914601802825928, "num_tokens": 401883186.0, "step": 10530 }, { "epoch": 1.339651443836662, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.172003746032715, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8523082733154297, "num_tokens": 401921149.0, "step": 10531 }, { "epoch": 1.3397786541152525, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.09328556060791, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.862434446811676, "num_tokens": 401961813.0, "step": 10532 }, { "epoch": 1.339905864393843, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.221564292907715, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8511884808540344, "num_tokens": 402001212.0, "step": 10533 }, { "epoch": 1.3400330746724336, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.143113136291504, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8619281053543091, "num_tokens": 402044648.0, "step": 10534 }, { "epoch": 1.3401602849510241, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.17172622680664, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8708919286727905, "num_tokens": 402087391.0, "step": 10535 }, { "epoch": 1.3402874952296147, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.186027526855469, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8887649774551392, "num_tokens": 402119598.0, "step": 10536 }, { "epoch": 1.3404147055082052, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.153738975524902, "learning_rate": 1e-06, "loss": 0.4945, "mean_token_accuracy": 0.8437484502792358, "num_tokens": 402153628.0, "step": 10537 }, { "epoch": 1.3405419157867955, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168259620666504, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8843740224838257, "num_tokens": 402190876.0, "step": 10538 }, { "epoch": 1.340669126065386, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.176678657531738, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8766699433326721, "num_tokens": 402227168.0, "step": 10539 }, { "epoch": 1.3407963363439765, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.182514190673828, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8771354556083679, "num_tokens": 402262731.0, "step": 10540 }, { "epoch": 1.340923546622567, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.232300758361816, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8693341016769409, "num_tokens": 402299992.0, "step": 10541 }, { "epoch": 1.3410507569011576, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.158358573913574, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8610391616821289, "num_tokens": 402339129.0, "step": 10542 }, { "epoch": 1.3411779671797481, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.163215637207031, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8513249754905701, "num_tokens": 402375380.0, "step": 10543 }, { "epoch": 1.3413051774583387, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.15786075592041, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8564267754554749, "num_tokens": 402412159.0, "step": 10544 }, { "epoch": 1.3414323877369292, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.271780967712402, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8642082214355469, "num_tokens": 402446676.0, "step": 10545 }, { "epoch": 1.3415595980155197, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.120333671569824, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8689647316932678, "num_tokens": 402490741.0, "step": 10546 }, { "epoch": 1.3416868082941102, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.153610229492188, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8769921064376831, "num_tokens": 402527398.0, "step": 10547 }, { "epoch": 1.3418140185727006, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.211008071899414, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.867844820022583, "num_tokens": 402565914.0, "step": 10548 }, { "epoch": 1.341941228851291, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.199223518371582, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8715931177139282, "num_tokens": 402600671.0, "step": 10549 }, { "epoch": 1.3420684391298816, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.187519073486328, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8808091878890991, "num_tokens": 402637381.0, "step": 10550 }, { "epoch": 1.3421956494084721, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.153249740600586, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8712307810783386, "num_tokens": 402671190.0, "step": 10551 }, { "epoch": 1.3423228596870627, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.153848648071289, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8680012226104736, "num_tokens": 402721518.0, "step": 10552 }, { "epoch": 1.3424500699656532, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.087531089782715, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8686462044715881, "num_tokens": 402758949.0, "step": 10553 }, { "epoch": 1.3425772802442437, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.22114372253418, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8624594211578369, "num_tokens": 402794342.0, "step": 10554 }, { "epoch": 1.3427044905228342, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.122532844543457, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8799220323562622, "num_tokens": 402831031.0, "step": 10555 }, { "epoch": 1.3428317008014248, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.144791603088379, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8786411285400391, "num_tokens": 402871486.0, "step": 10556 }, { "epoch": 1.3429589110800153, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.124706268310547, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8804589509963989, "num_tokens": 402916024.0, "step": 10557 }, { "epoch": 1.3430861213586058, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.226666450500488, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8502904176712036, "num_tokens": 402952116.0, "step": 10558 }, { "epoch": 1.3432133316371964, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.179947853088379, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.867233157157898, "num_tokens": 402998064.0, "step": 10559 }, { "epoch": 1.343340541915787, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.155287742614746, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8732298612594604, "num_tokens": 403041763.0, "step": 10560 }, { "epoch": 1.3434677521943774, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.213010787963867, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8798168897628784, "num_tokens": 403080956.0, "step": 10561 }, { "epoch": 1.343594962472968, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.296857833862305, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8776692748069763, "num_tokens": 403117200.0, "step": 10562 }, { "epoch": 1.3437221727515583, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.210371971130371, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8688426613807678, "num_tokens": 403154112.0, "step": 10563 }, { "epoch": 1.3438493830301488, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.163512229919434, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.86668860912323, "num_tokens": 403188265.0, "step": 10564 }, { "epoch": 1.3439765933087393, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.168451309204102, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8467127084732056, "num_tokens": 403231236.0, "step": 10565 }, { "epoch": 1.3441038035873298, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.14035701751709, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.870285153388977, "num_tokens": 403268642.0, "step": 10566 }, { "epoch": 1.3442310138659204, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.179550170898438, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8761087656021118, "num_tokens": 403307679.0, "step": 10567 }, { "epoch": 1.344358224144511, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.167705535888672, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8738348484039307, "num_tokens": 403346395.0, "step": 10568 }, { "epoch": 1.3444854344231014, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.192362785339355, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.878004789352417, "num_tokens": 403383001.0, "step": 10569 }, { "epoch": 1.344612644701692, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.15103530883789, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8606415390968323, "num_tokens": 403426090.0, "step": 10570 }, { "epoch": 1.3447398549802825, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.185012817382812, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8621965646743774, "num_tokens": 403470231.0, "step": 10571 }, { "epoch": 1.3448670652588728, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.216524124145508, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8738505244255066, "num_tokens": 403508677.0, "step": 10572 }, { "epoch": 1.3449942755374633, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.221692085266113, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.850103497505188, "num_tokens": 403545175.0, "step": 10573 }, { "epoch": 1.3451214858160538, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.213748931884766, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8819966316223145, "num_tokens": 403583321.0, "step": 10574 }, { "epoch": 1.3452486960946444, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.162633895874023, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8619186878204346, "num_tokens": 403618547.0, "step": 10575 }, { "epoch": 1.345375906373235, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.214859008789062, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8717961311340332, "num_tokens": 403657552.0, "step": 10576 }, { "epoch": 1.3455031166518254, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.214311599731445, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.871117889881134, "num_tokens": 403697021.0, "step": 10577 }, { "epoch": 1.345630326930416, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.21055793762207, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.858980655670166, "num_tokens": 403736480.0, "step": 10578 }, { "epoch": 1.3457575372090065, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.227785110473633, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8747936487197876, "num_tokens": 403772056.0, "step": 10579 }, { "epoch": 1.345884747487597, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.220078468322754, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8631356954574585, "num_tokens": 403813646.0, "step": 10580 }, { "epoch": 1.3460119577661875, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.174920082092285, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8716977834701538, "num_tokens": 403853904.0, "step": 10581 }, { "epoch": 1.346139168044778, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.150588989257812, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8527886867523193, "num_tokens": 403898528.0, "step": 10582 }, { "epoch": 1.3462663783233686, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.20145320892334, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8545621633529663, "num_tokens": 403929518.0, "step": 10583 }, { "epoch": 1.3463935886019591, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.174257278442383, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8702585101127625, "num_tokens": 403970323.0, "step": 10584 }, { "epoch": 1.3465207988805497, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.154747009277344, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.871756374835968, "num_tokens": 404012499.0, "step": 10585 }, { "epoch": 1.3466480091591402, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.159049034118652, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8639877438545227, "num_tokens": 404055370.0, "step": 10586 }, { "epoch": 1.3467752194377305, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.174543380737305, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8682467937469482, "num_tokens": 404092498.0, "step": 10587 }, { "epoch": 1.346902429716321, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168447494506836, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.87440425157547, "num_tokens": 404132466.0, "step": 10588 }, { "epoch": 1.3470296399949115, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.158082962036133, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8677838444709778, "num_tokens": 404171913.0, "step": 10589 }, { "epoch": 1.347156850273502, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.166936874389648, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8816622495651245, "num_tokens": 404205817.0, "step": 10590 }, { "epoch": 1.3472840605520926, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.192544937133789, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.871172308921814, "num_tokens": 404243922.0, "step": 10591 }, { "epoch": 1.3474112708306831, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.096586227416992, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8713979721069336, "num_tokens": 404283392.0, "step": 10592 }, { "epoch": 1.3475384811092737, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.201484680175781, "learning_rate": 1e-06, "loss": 0.4893, "mean_token_accuracy": 0.8430618047714233, "num_tokens": 404323044.0, "step": 10593 }, { "epoch": 1.3476656913878642, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.133974075317383, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8748949766159058, "num_tokens": 404359761.0, "step": 10594 }, { "epoch": 1.3477929016664547, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.203584671020508, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8555631637573242, "num_tokens": 404395354.0, "step": 10595 }, { "epoch": 1.3479201119450452, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.205133438110352, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8517436981201172, "num_tokens": 404432052.0, "step": 10596 }, { "epoch": 1.3480473222236355, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.21288776397705, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8612910509109497, "num_tokens": 404472689.0, "step": 10597 }, { "epoch": 1.348174532502226, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.148760795593262, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8530951142311096, "num_tokens": 404515916.0, "step": 10598 }, { "epoch": 1.3483017427808166, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.175357818603516, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8680397272109985, "num_tokens": 404557399.0, "step": 10599 }, { "epoch": 1.3484289530594071, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.229745864868164, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8643380403518677, "num_tokens": 404592573.0, "step": 10600 }, { "epoch": 1.3485561633379977, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.226106643676758, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8653889894485474, "num_tokens": 404635266.0, "step": 10601 }, { "epoch": 1.3486833736165882, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.191225051879883, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8774809241294861, "num_tokens": 404673004.0, "step": 10602 }, { "epoch": 1.3488105838951787, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.248416900634766, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8540865778923035, "num_tokens": 404705153.0, "step": 10603 }, { "epoch": 1.3489377941737692, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.260213851928711, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8597190380096436, "num_tokens": 404737068.0, "step": 10604 }, { "epoch": 1.3490650044523598, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.255731582641602, "learning_rate": 1e-06, "loss": 0.4866, "mean_token_accuracy": 0.8463016748428345, "num_tokens": 404773680.0, "step": 10605 }, { "epoch": 1.3491922147309503, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.150996208190918, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8735356330871582, "num_tokens": 404809937.0, "step": 10606 }, { "epoch": 1.3493194250095408, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.240487098693848, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8589949607849121, "num_tokens": 404847146.0, "step": 10607 }, { "epoch": 1.3494466352881314, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.21141529083252, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8691598176956177, "num_tokens": 404882952.0, "step": 10608 }, { "epoch": 1.3495738455667219, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.141838073730469, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8771824240684509, "num_tokens": 404925347.0, "step": 10609 }, { "epoch": 1.3497010558453124, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.132475852966309, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8541774749755859, "num_tokens": 404963411.0, "step": 10610 }, { "epoch": 1.349828266123903, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.193571090698242, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8663074374198914, "num_tokens": 404997522.0, "step": 10611 }, { "epoch": 1.3499554764024932, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.163240432739258, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8661488890647888, "num_tokens": 405034049.0, "step": 10612 }, { "epoch": 1.3500826866810838, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.170515060424805, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8646969199180603, "num_tokens": 405074985.0, "step": 10613 }, { "epoch": 1.3502098969596743, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.164566993713379, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8824669718742371, "num_tokens": 405116133.0, "step": 10614 }, { "epoch": 1.3503371072382648, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.21246337890625, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8654959201812744, "num_tokens": 405155572.0, "step": 10615 }, { "epoch": 1.3504643175168554, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.199773788452148, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8700546622276306, "num_tokens": 405194700.0, "step": 10616 }, { "epoch": 1.350591527795446, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.211118698120117, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.86358642578125, "num_tokens": 405233582.0, "step": 10617 }, { "epoch": 1.3507187380740364, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168313980102539, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8803740739822388, "num_tokens": 405274987.0, "step": 10618 }, { "epoch": 1.350845948352627, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.186596870422363, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.866034984588623, "num_tokens": 405308151.0, "step": 10619 }, { "epoch": 1.3509731586312175, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.138781547546387, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8675562143325806, "num_tokens": 405343939.0, "step": 10620 }, { "epoch": 1.3511003689098078, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.15943431854248, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8665152788162231, "num_tokens": 405384211.0, "step": 10621 }, { "epoch": 1.3512275791883983, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.1984224319458, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8748739957809448, "num_tokens": 405419971.0, "step": 10622 }, { "epoch": 1.3513547894669888, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.245956420898438, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8621417880058289, "num_tokens": 405451322.0, "step": 10623 }, { "epoch": 1.3514819997455794, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.203333854675293, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.859456479549408, "num_tokens": 405496101.0, "step": 10624 }, { "epoch": 1.35160921002417, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.20407772064209, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8516446352005005, "num_tokens": 405530034.0, "step": 10625 }, { "epoch": 1.3517364203027604, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.167607307434082, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8789845705032349, "num_tokens": 405570893.0, "step": 10626 }, { "epoch": 1.351863630581351, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.207087516784668, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8545940518379211, "num_tokens": 405610120.0, "step": 10627 }, { "epoch": 1.3519908408599415, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.191093444824219, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8669910430908203, "num_tokens": 405651878.0, "step": 10628 }, { "epoch": 1.352118051138532, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.197729110717773, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8722225427627563, "num_tokens": 405694578.0, "step": 10629 }, { "epoch": 1.3522452614171225, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.153867721557617, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.87394118309021, "num_tokens": 405734128.0, "step": 10630 }, { "epoch": 1.352372471695713, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.215493202209473, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8731110095977783, "num_tokens": 405776588.0, "step": 10631 }, { "epoch": 1.3524996819743036, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168140411376953, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8638800382614136, "num_tokens": 405810867.0, "step": 10632 }, { "epoch": 1.3526268922528941, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.232705116271973, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8855482339859009, "num_tokens": 405848884.0, "step": 10633 }, { "epoch": 1.3527541025314846, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.206856727600098, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8874796628952026, "num_tokens": 405888759.0, "step": 10634 }, { "epoch": 1.3528813128100752, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.160622596740723, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8816623091697693, "num_tokens": 405925686.0, "step": 10635 }, { "epoch": 1.3530085230886655, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.252464294433594, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8811517357826233, "num_tokens": 405960836.0, "step": 10636 }, { "epoch": 1.353135733367256, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.168380737304688, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8787181377410889, "num_tokens": 406000314.0, "step": 10637 }, { "epoch": 1.3532629436458465, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.14608097076416, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8781536817550659, "num_tokens": 406039468.0, "step": 10638 }, { "epoch": 1.353390153924437, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.190878868103027, "learning_rate": 1e-06, "loss": 0.3416, "mean_token_accuracy": 0.8923976421356201, "num_tokens": 406076828.0, "step": 10639 }, { "epoch": 1.3535173642030276, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.164135932922363, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8674973249435425, "num_tokens": 406116504.0, "step": 10640 }, { "epoch": 1.3536445744816181, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.207518577575684, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8676864504814148, "num_tokens": 406155184.0, "step": 10641 }, { "epoch": 1.3537717847602087, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.142496109008789, "learning_rate": 1e-06, "loss": 0.4907, "mean_token_accuracy": 0.849396288394928, "num_tokens": 406195411.0, "step": 10642 }, { "epoch": 1.3538989950387992, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.158831596374512, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8587086200714111, "num_tokens": 406230159.0, "step": 10643 }, { "epoch": 1.3540262053173897, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.192112922668457, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.857512354850769, "num_tokens": 406264541.0, "step": 10644 }, { "epoch": 1.3541534155959802, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.170730590820312, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8733656406402588, "num_tokens": 406301869.0, "step": 10645 }, { "epoch": 1.3542806258745705, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.173819541931152, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8692651987075806, "num_tokens": 406338819.0, "step": 10646 }, { "epoch": 1.354407836153161, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.183650970458984, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8697882890701294, "num_tokens": 406375914.0, "step": 10647 }, { "epoch": 1.3545350464317516, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.211775779724121, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8729749917984009, "num_tokens": 406418543.0, "step": 10648 }, { "epoch": 1.3546622567103421, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.173309326171875, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.866258978843689, "num_tokens": 406455731.0, "step": 10649 }, { "epoch": 1.3547894669889327, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.230319023132324, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8678506016731262, "num_tokens": 406496266.0, "step": 10650 }, { "epoch": 1.3549166772675232, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.163741111755371, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8636833429336548, "num_tokens": 406541676.0, "step": 10651 }, { "epoch": 1.3550438875461137, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.2413911819458, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8643282055854797, "num_tokens": 406578997.0, "step": 10652 }, { "epoch": 1.3551710978247042, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.201728820800781, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8727734684944153, "num_tokens": 406619404.0, "step": 10653 }, { "epoch": 1.3552983081032948, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.201777458190918, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8855676651000977, "num_tokens": 406656092.0, "step": 10654 }, { "epoch": 1.3554255183818853, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.14658260345459, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8514646291732788, "num_tokens": 406701600.0, "step": 10655 }, { "epoch": 1.3555527286604758, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.227685928344727, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8675045371055603, "num_tokens": 406739649.0, "step": 10656 }, { "epoch": 1.3556799389390664, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.206764221191406, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8877478837966919, "num_tokens": 406774962.0, "step": 10657 }, { "epoch": 1.3558071492176569, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.132770538330078, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8651012778282166, "num_tokens": 406818047.0, "step": 10658 }, { "epoch": 1.3559343594962474, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.199715614318848, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8618264198303223, "num_tokens": 406859464.0, "step": 10659 }, { "epoch": 1.356061569774838, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.202930450439453, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8667854070663452, "num_tokens": 406896444.0, "step": 10660 }, { "epoch": 1.3561887800534282, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.214298248291016, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.862678587436676, "num_tokens": 406934623.0, "step": 10661 }, { "epoch": 1.3563159903320188, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.190923690795898, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8609713912010193, "num_tokens": 406974290.0, "step": 10662 }, { "epoch": 1.3564432006106093, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.122973442077637, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8869536519050598, "num_tokens": 407011279.0, "step": 10663 }, { "epoch": 1.3565704108891998, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.157209396362305, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8578149080276489, "num_tokens": 407054733.0, "step": 10664 }, { "epoch": 1.3566976211677904, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.125967979431152, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8691326379776001, "num_tokens": 407098410.0, "step": 10665 }, { "epoch": 1.3568248314463809, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.190563201904297, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8624832630157471, "num_tokens": 407134231.0, "step": 10666 }, { "epoch": 1.3569520417249714, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.17691707611084, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.853705108165741, "num_tokens": 407167231.0, "step": 10667 }, { "epoch": 1.357079252003562, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.193532943725586, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8712056279182434, "num_tokens": 407206604.0, "step": 10668 }, { "epoch": 1.3572064622821525, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.16723346710205, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8703720569610596, "num_tokens": 407248247.0, "step": 10669 }, { "epoch": 1.3573336725607428, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.141593933105469, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8781124353408813, "num_tokens": 407282870.0, "step": 10670 }, { "epoch": 1.3574608828393333, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.195249557495117, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8750373125076294, "num_tokens": 407318704.0, "step": 10671 }, { "epoch": 1.3575880931179238, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.16308879852295, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.870532214641571, "num_tokens": 407360003.0, "step": 10672 }, { "epoch": 1.3577153033965144, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.123214721679688, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8625435829162598, "num_tokens": 407399473.0, "step": 10673 }, { "epoch": 1.357842513675105, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.1651029586792, "learning_rate": 1e-06, "loss": 0.5041, "mean_token_accuracy": 0.8380991816520691, "num_tokens": 407440576.0, "step": 10674 }, { "epoch": 1.3579697239536954, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.207788467407227, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8651466369628906, "num_tokens": 407481057.0, "step": 10675 }, { "epoch": 1.358096934232286, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.17489242553711, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8803498148918152, "num_tokens": 407517104.0, "step": 10676 }, { "epoch": 1.3582241445108765, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.238052368164062, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8757021427154541, "num_tokens": 407552604.0, "step": 10677 }, { "epoch": 1.358351354789467, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.176392555236816, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8823943138122559, "num_tokens": 407588282.0, "step": 10678 }, { "epoch": 1.3584785650680575, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.180806159973145, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8635197877883911, "num_tokens": 407619837.0, "step": 10679 }, { "epoch": 1.358605775346648, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.19892692565918, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.865458071231842, "num_tokens": 407656939.0, "step": 10680 }, { "epoch": 1.3587329856252386, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.206192970275879, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8586713671684265, "num_tokens": 407702756.0, "step": 10681 }, { "epoch": 1.3588601959038291, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.185698509216309, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8768607378005981, "num_tokens": 407736142.0, "step": 10682 }, { "epoch": 1.3589874061824196, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.257796287536621, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.864886462688446, "num_tokens": 407773184.0, "step": 10683 }, { "epoch": 1.3591146164610102, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.232966423034668, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8537781834602356, "num_tokens": 407806543.0, "step": 10684 }, { "epoch": 1.3592418267396005, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.249972343444824, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8670961856842041, "num_tokens": 407841371.0, "step": 10685 }, { "epoch": 1.359369037018191, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.191978454589844, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8707436323165894, "num_tokens": 407879862.0, "step": 10686 }, { "epoch": 1.3594962472967815, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.286198616027832, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8623465299606323, "num_tokens": 407917947.0, "step": 10687 }, { "epoch": 1.359623457575372, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.242385864257812, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.875541090965271, "num_tokens": 407953987.0, "step": 10688 }, { "epoch": 1.3597506678539626, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.221988677978516, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8521044254302979, "num_tokens": 407993166.0, "step": 10689 }, { "epoch": 1.3598778781325531, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.215935707092285, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8769397735595703, "num_tokens": 408037063.0, "step": 10690 }, { "epoch": 1.3600050884111436, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.215100288391113, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8712227940559387, "num_tokens": 408078526.0, "step": 10691 }, { "epoch": 1.3601322986897342, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.229340553283691, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8756105303764343, "num_tokens": 408116002.0, "step": 10692 }, { "epoch": 1.3602595089683247, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.196749687194824, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8639785051345825, "num_tokens": 408155885.0, "step": 10693 }, { "epoch": 1.3603867192469152, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.283539772033691, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8736634850502014, "num_tokens": 408190085.0, "step": 10694 }, { "epoch": 1.3605139295255055, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.233424186706543, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8704334497451782, "num_tokens": 408227289.0, "step": 10695 }, { "epoch": 1.360641139804096, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.229926109313965, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8767493963241577, "num_tokens": 408271853.0, "step": 10696 }, { "epoch": 1.3607683500826866, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.211249351501465, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8806270360946655, "num_tokens": 408311133.0, "step": 10697 }, { "epoch": 1.3608955603612771, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.266580581665039, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8699946999549866, "num_tokens": 408342775.0, "step": 10698 }, { "epoch": 1.3610227706398677, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.178948402404785, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8807976841926575, "num_tokens": 408379063.0, "step": 10699 }, { "epoch": 1.3611499809184582, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.274056434631348, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8560101985931396, "num_tokens": 408415856.0, "step": 10700 }, { "epoch": 1.3612771911970487, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.241339683532715, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8726877570152283, "num_tokens": 408451522.0, "step": 10701 }, { "epoch": 1.3614044014756392, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.147224426269531, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8735470771789551, "num_tokens": 408494605.0, "step": 10702 }, { "epoch": 1.3615316117542298, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.231123924255371, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8672192692756653, "num_tokens": 408536373.0, "step": 10703 }, { "epoch": 1.3616588220328203, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.178376197814941, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8563786745071411, "num_tokens": 408574724.0, "step": 10704 }, { "epoch": 1.3617860323114108, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.26104736328125, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8609543442726135, "num_tokens": 408615057.0, "step": 10705 }, { "epoch": 1.3619132425900013, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.21969985961914, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.85602205991745, "num_tokens": 408651324.0, "step": 10706 }, { "epoch": 1.3620404528685919, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.249103546142578, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8746249675750732, "num_tokens": 408690533.0, "step": 10707 }, { "epoch": 1.3621676631471824, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.186578750610352, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8610550165176392, "num_tokens": 408725970.0, "step": 10708 }, { "epoch": 1.362294873425773, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.210189819335938, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8765912055969238, "num_tokens": 408766550.0, "step": 10709 }, { "epoch": 1.3624220837043632, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.250975608825684, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8725558519363403, "num_tokens": 408804320.0, "step": 10710 }, { "epoch": 1.3625492939829538, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.231101036071777, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8705332279205322, "num_tokens": 408840902.0, "step": 10711 }, { "epoch": 1.3626765042615443, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.23460578918457, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8667057752609253, "num_tokens": 408879346.0, "step": 10712 }, { "epoch": 1.3628037145401348, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.252594947814941, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8806518316268921, "num_tokens": 408919336.0, "step": 10713 }, { "epoch": 1.3629309248187254, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.118304252624512, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8639166355133057, "num_tokens": 408961547.0, "step": 10714 }, { "epoch": 1.3630581350973159, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.246391296386719, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8836609125137329, "num_tokens": 409003573.0, "step": 10715 }, { "epoch": 1.3631853453759064, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.186979293823242, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8620781302452087, "num_tokens": 409045560.0, "step": 10716 }, { "epoch": 1.363312555654497, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.262656211853027, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8726218938827515, "num_tokens": 409087245.0, "step": 10717 }, { "epoch": 1.3634397659330875, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.201482772827148, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8773547410964966, "num_tokens": 409134177.0, "step": 10718 }, { "epoch": 1.3635669762116778, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.244139671325684, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8551619648933411, "num_tokens": 409169495.0, "step": 10719 }, { "epoch": 1.3636941864902683, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.165979385375977, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.8442242741584778, "num_tokens": 409207781.0, "step": 10720 }, { "epoch": 1.3638213967688588, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.2606782913208, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8680245280265808, "num_tokens": 409245411.0, "step": 10721 }, { "epoch": 1.3639486070474494, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.246575355529785, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8765230178833008, "num_tokens": 409281757.0, "step": 10722 }, { "epoch": 1.3640758173260399, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.234901428222656, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8571043014526367, "num_tokens": 409321317.0, "step": 10723 }, { "epoch": 1.3642030276046304, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.155539512634277, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8750335574150085, "num_tokens": 409358919.0, "step": 10724 }, { "epoch": 1.364330237883221, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.266983032226562, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8597821593284607, "num_tokens": 409394355.0, "step": 10725 }, { "epoch": 1.3644574481618115, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.227378845214844, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8777942657470703, "num_tokens": 409433781.0, "step": 10726 }, { "epoch": 1.364584658440402, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.17934513092041, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8581637740135193, "num_tokens": 409471728.0, "step": 10727 }, { "epoch": 1.3647118687189925, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.237716674804688, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8871061205863953, "num_tokens": 409502294.0, "step": 10728 }, { "epoch": 1.364839078997583, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.161530494689941, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8701205849647522, "num_tokens": 409545878.0, "step": 10729 }, { "epoch": 1.3649662892761736, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.222628593444824, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8700745105743408, "num_tokens": 409584753.0, "step": 10730 }, { "epoch": 1.3650934995547641, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.233040809631348, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8731056451797485, "num_tokens": 409613725.0, "step": 10731 }, { "epoch": 1.3652207098333546, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.217172622680664, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8621139526367188, "num_tokens": 409652443.0, "step": 10732 }, { "epoch": 1.3653479201119452, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.268542289733887, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8751897811889648, "num_tokens": 409690683.0, "step": 10733 }, { "epoch": 1.3654751303905355, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.179478645324707, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8687986135482788, "num_tokens": 409727502.0, "step": 10734 }, { "epoch": 1.365602340669126, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.239882469177246, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.864117443561554, "num_tokens": 409764592.0, "step": 10735 }, { "epoch": 1.3657295509477165, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.157176971435547, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8465392589569092, "num_tokens": 409809645.0, "step": 10736 }, { "epoch": 1.365856761226307, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.220588684082031, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.868270218372345, "num_tokens": 409848048.0, "step": 10737 }, { "epoch": 1.3659839715048976, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.184869766235352, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8617634177207947, "num_tokens": 409884176.0, "step": 10738 }, { "epoch": 1.3661111817834881, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.162946701049805, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8704165816307068, "num_tokens": 409922630.0, "step": 10739 }, { "epoch": 1.3662383920620786, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.195746421813965, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8745143413543701, "num_tokens": 409961876.0, "step": 10740 }, { "epoch": 1.3663656023406692, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.139230728149414, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.876889169216156, "num_tokens": 409995432.0, "step": 10741 }, { "epoch": 1.3664928126192597, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.193512916564941, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8667334318161011, "num_tokens": 410034876.0, "step": 10742 }, { "epoch": 1.3666200228978502, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.18466854095459, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8718730211257935, "num_tokens": 410079190.0, "step": 10743 }, { "epoch": 1.3667472331764405, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.173982620239258, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8715499639511108, "num_tokens": 410112475.0, "step": 10744 }, { "epoch": 1.366874443455031, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.188216209411621, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8829803466796875, "num_tokens": 410154583.0, "step": 10745 }, { "epoch": 1.3670016537336216, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.16578197479248, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8688074946403503, "num_tokens": 410191631.0, "step": 10746 }, { "epoch": 1.3671288640122121, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.209113121032715, "learning_rate": 1e-06, "loss": 0.4672, "mean_token_accuracy": 0.8458677530288696, "num_tokens": 410232435.0, "step": 10747 }, { "epoch": 1.3672560742908026, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.171152114868164, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8735095262527466, "num_tokens": 410272940.0, "step": 10748 }, { "epoch": 1.3673832845693932, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.186318397521973, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8643147945404053, "num_tokens": 410312807.0, "step": 10749 }, { "epoch": 1.3675104948479837, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.19582748413086, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8739367127418518, "num_tokens": 410352433.0, "step": 10750 }, { "epoch": 1.3676377051265742, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.192835807800293, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8696091771125793, "num_tokens": 410394843.0, "step": 10751 }, { "epoch": 1.3677649154051648, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.197400093078613, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8702350854873657, "num_tokens": 410439353.0, "step": 10752 }, { "epoch": 1.3678921256837553, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.188634872436523, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.846513032913208, "num_tokens": 410485789.0, "step": 10753 }, { "epoch": 1.3680193359623458, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.149168968200684, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8530908226966858, "num_tokens": 410523385.0, "step": 10754 }, { "epoch": 1.3681465462409363, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.262824058532715, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8629611134529114, "num_tokens": 410553340.0, "step": 10755 }, { "epoch": 1.3682737565195269, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.198598861694336, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8765856027603149, "num_tokens": 410590041.0, "step": 10756 }, { "epoch": 1.3684009667981174, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.201802253723145, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8707625865936279, "num_tokens": 410627703.0, "step": 10757 }, { "epoch": 1.368528177076708, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.16543197631836, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8613319396972656, "num_tokens": 410662555.0, "step": 10758 }, { "epoch": 1.3686553873552982, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.23121166229248, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8505880832672119, "num_tokens": 410706026.0, "step": 10759 }, { "epoch": 1.3687825976338888, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.253385543823242, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8655456900596619, "num_tokens": 410743275.0, "step": 10760 }, { "epoch": 1.3689098079124793, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.166152000427246, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8772412538528442, "num_tokens": 410781220.0, "step": 10761 }, { "epoch": 1.3690370181910698, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.274051666259766, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8763057589530945, "num_tokens": 410820014.0, "step": 10762 }, { "epoch": 1.3691642284696603, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.200482368469238, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8640251159667969, "num_tokens": 410858721.0, "step": 10763 }, { "epoch": 1.3692914387482509, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.257621765136719, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8732424974441528, "num_tokens": 410898527.0, "step": 10764 }, { "epoch": 1.3694186490268414, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.230069160461426, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8694297075271606, "num_tokens": 410938426.0, "step": 10765 }, { "epoch": 1.369545859305432, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.245316505432129, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8737677931785583, "num_tokens": 410969940.0, "step": 10766 }, { "epoch": 1.3696730695840225, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.29978084564209, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8530384302139282, "num_tokens": 411008859.0, "step": 10767 }, { "epoch": 1.3698002798626128, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.209370613098145, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8814717531204224, "num_tokens": 411043516.0, "step": 10768 }, { "epoch": 1.3699274901412033, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.235688209533691, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8612066507339478, "num_tokens": 411079092.0, "step": 10769 }, { "epoch": 1.3700547004197938, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.233574867248535, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8725054860115051, "num_tokens": 411117360.0, "step": 10770 }, { "epoch": 1.3701819106983844, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.249810218811035, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8581352829933167, "num_tokens": 411153615.0, "step": 10771 }, { "epoch": 1.3703091209769749, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.23369312286377, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8707208037376404, "num_tokens": 411193571.0, "step": 10772 }, { "epoch": 1.3704363312555654, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.239456176757812, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8505196571350098, "num_tokens": 411230427.0, "step": 10773 }, { "epoch": 1.370563541534156, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.182966232299805, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8663625121116638, "num_tokens": 411271251.0, "step": 10774 }, { "epoch": 1.3706907518127465, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.27985668182373, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8701044917106628, "num_tokens": 411304906.0, "step": 10775 }, { "epoch": 1.370817962091337, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.302128791809082, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8849213719367981, "num_tokens": 411341417.0, "step": 10776 }, { "epoch": 1.3709451723699275, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.183470726013184, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8691641092300415, "num_tokens": 411382118.0, "step": 10777 }, { "epoch": 1.371072382648518, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.256836891174316, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8569297790527344, "num_tokens": 411426702.0, "step": 10778 }, { "epoch": 1.3711995929271086, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.27741527557373, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8679594397544861, "num_tokens": 411461689.0, "step": 10779 }, { "epoch": 1.371326803205699, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.257842063903809, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8681253790855408, "num_tokens": 411499613.0, "step": 10780 }, { "epoch": 1.3714540134842896, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.236635208129883, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8675409555435181, "num_tokens": 411538422.0, "step": 10781 }, { "epoch": 1.3715812237628802, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.239275932312012, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8812852501869202, "num_tokens": 411572401.0, "step": 10782 }, { "epoch": 1.3717084340414705, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.301755905151367, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8749580383300781, "num_tokens": 411611268.0, "step": 10783 }, { "epoch": 1.371835644320061, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.256319999694824, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8687669038772583, "num_tokens": 411648837.0, "step": 10784 }, { "epoch": 1.3719628545986515, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.229219436645508, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8718063831329346, "num_tokens": 411690327.0, "step": 10785 }, { "epoch": 1.372090064877242, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.254366874694824, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.867024302482605, "num_tokens": 411729665.0, "step": 10786 }, { "epoch": 1.3722172751558326, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.255101203918457, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8820456266403198, "num_tokens": 411764716.0, "step": 10787 }, { "epoch": 1.3723444854344231, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.204605102539062, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8692944049835205, "num_tokens": 411805560.0, "step": 10788 }, { "epoch": 1.3724716957130136, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.22461986541748, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8730496764183044, "num_tokens": 411841151.0, "step": 10789 }, { "epoch": 1.3725989059916042, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.258247375488281, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8755226135253906, "num_tokens": 411875088.0, "step": 10790 }, { "epoch": 1.3727261162701947, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.191839218139648, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8691806793212891, "num_tokens": 411914554.0, "step": 10791 }, { "epoch": 1.3728533265487852, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.219886779785156, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.869051456451416, "num_tokens": 411952237.0, "step": 10792 }, { "epoch": 1.3729805368273755, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.222319602966309, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8629765510559082, "num_tokens": 411989676.0, "step": 10793 }, { "epoch": 1.373107747105966, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.231538772583008, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8812916278839111, "num_tokens": 412025804.0, "step": 10794 }, { "epoch": 1.3732349573845566, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.19680118560791, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8847923278808594, "num_tokens": 412060085.0, "step": 10795 }, { "epoch": 1.3733621676631471, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.267849922180176, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.87503981590271, "num_tokens": 412100750.0, "step": 10796 }, { "epoch": 1.3734893779417376, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.257226943969727, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8713321685791016, "num_tokens": 412145829.0, "step": 10797 }, { "epoch": 1.3736165882203282, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.226359367370605, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8737248778343201, "num_tokens": 412187607.0, "step": 10798 }, { "epoch": 1.3737437984989187, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.303450584411621, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8558834791183472, "num_tokens": 412227214.0, "step": 10799 }, { "epoch": 1.3738710087775092, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.243718147277832, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8723707795143127, "num_tokens": 412261929.0, "step": 10800 }, { "epoch": 1.3739982190560998, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.266481399536133, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8650097846984863, "num_tokens": 412299441.0, "step": 10801 }, { "epoch": 1.3741254293346903, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.234944343566895, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8675535917282104, "num_tokens": 412339248.0, "step": 10802 }, { "epoch": 1.3742526396132808, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.292848587036133, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8681689500808716, "num_tokens": 412377224.0, "step": 10803 }, { "epoch": 1.3743798498918713, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.256141662597656, "learning_rate": 1e-06, "loss": 0.4793, "mean_token_accuracy": 0.8460538387298584, "num_tokens": 412415331.0, "step": 10804 }, { "epoch": 1.3745070601704619, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.239570617675781, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8685569763183594, "num_tokens": 412455604.0, "step": 10805 }, { "epoch": 1.3746342704490524, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.266510963439941, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.875522255897522, "num_tokens": 412489295.0, "step": 10806 }, { "epoch": 1.374761480727643, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.260677337646484, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8650026321411133, "num_tokens": 412525750.0, "step": 10807 }, { "epoch": 1.3748886910062332, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.23470687866211, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8775465488433838, "num_tokens": 412567163.0, "step": 10808 }, { "epoch": 1.3750159012848238, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.167261123657227, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8756858110427856, "num_tokens": 412598789.0, "step": 10809 }, { "epoch": 1.3751431115634143, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.225404739379883, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8718148469924927, "num_tokens": 412632588.0, "step": 10810 }, { "epoch": 1.3752703218420048, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.190656661987305, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8671275973320007, "num_tokens": 412676369.0, "step": 10811 }, { "epoch": 1.3753975321205953, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.183829307556152, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8834010362625122, "num_tokens": 412712674.0, "step": 10812 }, { "epoch": 1.3755247423991859, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.253668785095215, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8633322715759277, "num_tokens": 412753531.0, "step": 10813 }, { "epoch": 1.3756519526777764, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.259692192077637, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8656917214393616, "num_tokens": 412799450.0, "step": 10814 }, { "epoch": 1.375779162956367, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.152320861816406, "learning_rate": 1e-06, "loss": 0.3398, "mean_token_accuracy": 0.8903247117996216, "num_tokens": 412836858.0, "step": 10815 }, { "epoch": 1.3759063732349575, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.295915603637695, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8721074461936951, "num_tokens": 412871169.0, "step": 10816 }, { "epoch": 1.3760335835135478, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.233965873718262, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8736927509307861, "num_tokens": 412909281.0, "step": 10817 }, { "epoch": 1.3761607937921383, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.247739791870117, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.870182454586029, "num_tokens": 412950126.0, "step": 10818 }, { "epoch": 1.3762880040707288, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.199321746826172, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8570277094841003, "num_tokens": 412988644.0, "step": 10819 }, { "epoch": 1.3764152143493193, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.308075904846191, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8551158905029297, "num_tokens": 413028403.0, "step": 10820 }, { "epoch": 1.3765424246279099, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.278266906738281, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8602051734924316, "num_tokens": 413067429.0, "step": 10821 }, { "epoch": 1.3766696349065004, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.20510196685791, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8651703596115112, "num_tokens": 413101630.0, "step": 10822 }, { "epoch": 1.376796845185091, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.28875732421875, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8629090785980225, "num_tokens": 413147757.0, "step": 10823 }, { "epoch": 1.3769240554636815, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.25755786895752, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8714932203292847, "num_tokens": 413186803.0, "step": 10824 }, { "epoch": 1.377051265742272, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.250574111938477, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8629474639892578, "num_tokens": 413227201.0, "step": 10825 }, { "epoch": 1.3771784760208625, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.285857200622559, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8667498826980591, "num_tokens": 413270594.0, "step": 10826 }, { "epoch": 1.377305686299453, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.198314666748047, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8522054553031921, "num_tokens": 413307345.0, "step": 10827 }, { "epoch": 1.3774328965780436, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.305235862731934, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8732508420944214, "num_tokens": 413344310.0, "step": 10828 }, { "epoch": 1.377560106856634, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.174118995666504, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8875313997268677, "num_tokens": 413382047.0, "step": 10829 }, { "epoch": 1.3776873171352246, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.230507850646973, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8662035465240479, "num_tokens": 413424658.0, "step": 10830 }, { "epoch": 1.3778145274138152, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.238341331481934, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8770703673362732, "num_tokens": 413458386.0, "step": 10831 }, { "epoch": 1.3779417376924055, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.237448692321777, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8557751178741455, "num_tokens": 413492530.0, "step": 10832 }, { "epoch": 1.378068947970996, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.258298873901367, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8796502351760864, "num_tokens": 413528894.0, "step": 10833 }, { "epoch": 1.3781961582495865, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.271888732910156, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8649255633354187, "num_tokens": 413564765.0, "step": 10834 }, { "epoch": 1.378323368528177, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.221956253051758, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8416509628295898, "num_tokens": 413598122.0, "step": 10835 }, { "epoch": 1.3784505788067676, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.27308464050293, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8526005744934082, "num_tokens": 413634519.0, "step": 10836 }, { "epoch": 1.378577789085358, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.243098258972168, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8606823682785034, "num_tokens": 413674021.0, "step": 10837 }, { "epoch": 1.3787049993639486, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.342443466186523, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8748718500137329, "num_tokens": 413713804.0, "step": 10838 }, { "epoch": 1.3788322096425392, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.226339340209961, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8756625652313232, "num_tokens": 413755665.0, "step": 10839 }, { "epoch": 1.3789594199211297, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.321369171142578, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8677311539649963, "num_tokens": 413790872.0, "step": 10840 }, { "epoch": 1.3790866301997202, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.240737915039062, "learning_rate": 1e-06, "loss": 0.3461, "mean_token_accuracy": 0.8882278203964233, "num_tokens": 413830798.0, "step": 10841 }, { "epoch": 1.3792138404783105, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.28442096710205, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8731722831726074, "num_tokens": 413865960.0, "step": 10842 }, { "epoch": 1.379341050756901, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.259454727172852, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8788207173347473, "num_tokens": 413898222.0, "step": 10843 }, { "epoch": 1.3794682610354916, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.267810821533203, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8635866641998291, "num_tokens": 413931839.0, "step": 10844 }, { "epoch": 1.3795954713140821, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.317573547363281, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8594487309455872, "num_tokens": 413963815.0, "step": 10845 }, { "epoch": 1.3797226815926726, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.317501068115234, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8768774271011353, "num_tokens": 414002476.0, "step": 10846 }, { "epoch": 1.3798498918712632, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.268957138061523, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8637002110481262, "num_tokens": 414041121.0, "step": 10847 }, { "epoch": 1.3799771021498537, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.26608943939209, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.856879711151123, "num_tokens": 414080447.0, "step": 10848 }, { "epoch": 1.3801043124284442, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.299025535583496, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8595462441444397, "num_tokens": 414108925.0, "step": 10849 }, { "epoch": 1.3802315227070348, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.235820770263672, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8735220432281494, "num_tokens": 414154296.0, "step": 10850 }, { "epoch": 1.3803587329856253, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.2512788772583, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.856639564037323, "num_tokens": 414194612.0, "step": 10851 }, { "epoch": 1.3804859432642158, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.30305004119873, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8714245557785034, "num_tokens": 414229939.0, "step": 10852 }, { "epoch": 1.3806131535428063, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.242602348327637, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.865749716758728, "num_tokens": 414265193.0, "step": 10853 }, { "epoch": 1.3807403638213969, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.288055419921875, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8660883903503418, "num_tokens": 414301471.0, "step": 10854 }, { "epoch": 1.3808675740999874, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.208724021911621, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.8397598266601562, "num_tokens": 414338755.0, "step": 10855 }, { "epoch": 1.380994784378578, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.307689666748047, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8696860074996948, "num_tokens": 414379996.0, "step": 10856 }, { "epoch": 1.3811219946571682, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.19804573059082, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8720792531967163, "num_tokens": 414421905.0, "step": 10857 }, { "epoch": 1.3812492049357588, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.257139205932617, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8737492561340332, "num_tokens": 414465149.0, "step": 10858 }, { "epoch": 1.3813764152143493, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.163022994995117, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8780996799468994, "num_tokens": 414504086.0, "step": 10859 }, { "epoch": 1.3815036254929398, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.29502010345459, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8520547151565552, "num_tokens": 414545392.0, "step": 10860 }, { "epoch": 1.3816308357715303, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.261889457702637, "learning_rate": 1e-06, "loss": 0.514, "mean_token_accuracy": 0.8331523537635803, "num_tokens": 414586445.0, "step": 10861 }, { "epoch": 1.3817580460501209, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.32364273071289, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8774611949920654, "num_tokens": 414618467.0, "step": 10862 }, { "epoch": 1.3818852563287114, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.241421699523926, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8785581588745117, "num_tokens": 414651253.0, "step": 10863 }, { "epoch": 1.382012466607302, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.267096519470215, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8678717613220215, "num_tokens": 414692016.0, "step": 10864 }, { "epoch": 1.3821396768858925, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.33624267578125, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8735663890838623, "num_tokens": 414731840.0, "step": 10865 }, { "epoch": 1.3822668871644828, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.182497024536133, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.870517909526825, "num_tokens": 414767110.0, "step": 10866 }, { "epoch": 1.3823940974430733, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.295516967773438, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8744009137153625, "num_tokens": 414799340.0, "step": 10867 }, { "epoch": 1.3825213077216638, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.233553886413574, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8713295459747314, "num_tokens": 414840915.0, "step": 10868 }, { "epoch": 1.3826485180002543, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.227738380432129, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.857570230960846, "num_tokens": 414877298.0, "step": 10869 }, { "epoch": 1.3827757282788449, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.203699111938477, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8756037950515747, "num_tokens": 414916431.0, "step": 10870 }, { "epoch": 1.3829029385574354, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.263982772827148, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8668756484985352, "num_tokens": 414958148.0, "step": 10871 }, { "epoch": 1.383030148836026, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.237370491027832, "learning_rate": 1e-06, "loss": 0.3354, "mean_token_accuracy": 0.8932995796203613, "num_tokens": 414997008.0, "step": 10872 }, { "epoch": 1.3831573591146165, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.288857460021973, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8553814888000488, "num_tokens": 415042893.0, "step": 10873 }, { "epoch": 1.383284569393207, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.250107765197754, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8669446706771851, "num_tokens": 415077476.0, "step": 10874 }, { "epoch": 1.3834117796717975, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.247437477111816, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8632080554962158, "num_tokens": 415124329.0, "step": 10875 }, { "epoch": 1.383538989950388, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.274855613708496, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.879559338092804, "num_tokens": 415164101.0, "step": 10876 }, { "epoch": 1.3836662002289786, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.336953163146973, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8686926364898682, "num_tokens": 415203792.0, "step": 10877 }, { "epoch": 1.383793410507569, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.306289672851562, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8658396005630493, "num_tokens": 415233754.0, "step": 10878 }, { "epoch": 1.3839206207861596, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.29056453704834, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.876204252243042, "num_tokens": 415269567.0, "step": 10879 }, { "epoch": 1.3840478310647502, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.274151802062988, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8721844553947449, "num_tokens": 415312660.0, "step": 10880 }, { "epoch": 1.3841750413433405, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.26720905303955, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8788546323776245, "num_tokens": 415355377.0, "step": 10881 }, { "epoch": 1.384302251621931, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.309405326843262, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8799276351928711, "num_tokens": 415393109.0, "step": 10882 }, { "epoch": 1.3844294619005215, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.279122352600098, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8803982138633728, "num_tokens": 415428650.0, "step": 10883 }, { "epoch": 1.384556672179112, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.281603813171387, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.866174578666687, "num_tokens": 415467166.0, "step": 10884 }, { "epoch": 1.3846838824577026, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.291780471801758, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.881624162197113, "num_tokens": 415505633.0, "step": 10885 }, { "epoch": 1.384811092736293, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.315995216369629, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8682384490966797, "num_tokens": 415546254.0, "step": 10886 }, { "epoch": 1.3849383030148836, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.25539493560791, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8672388195991516, "num_tokens": 415584999.0, "step": 10887 }, { "epoch": 1.3850655132934742, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.250768661499023, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8709502220153809, "num_tokens": 415624608.0, "step": 10888 }, { "epoch": 1.3851927235720647, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.313018798828125, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.855597734451294, "num_tokens": 415663405.0, "step": 10889 }, { "epoch": 1.385319933850655, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.280184745788574, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8824891448020935, "num_tokens": 415698934.0, "step": 10890 }, { "epoch": 1.3854471441292455, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.234637260437012, "learning_rate": 1e-06, "loss": 0.497, "mean_token_accuracy": 0.8410828709602356, "num_tokens": 415736973.0, "step": 10891 }, { "epoch": 1.385574354407836, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.245408058166504, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8718817234039307, "num_tokens": 415778364.0, "step": 10892 }, { "epoch": 1.3857015646864266, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.249768257141113, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8847568035125732, "num_tokens": 415817502.0, "step": 10893 }, { "epoch": 1.385828774965017, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.304450035095215, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8641103506088257, "num_tokens": 415860364.0, "step": 10894 }, { "epoch": 1.3859559852436076, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.256837844848633, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8725525140762329, "num_tokens": 415896816.0, "step": 10895 }, { "epoch": 1.3860831955221982, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.263291358947754, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8516116142272949, "num_tokens": 415935173.0, "step": 10896 }, { "epoch": 1.3862104058007887, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.25090217590332, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.878536581993103, "num_tokens": 415977043.0, "step": 10897 }, { "epoch": 1.3863376160793792, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.25927734375, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8782267570495605, "num_tokens": 416014376.0, "step": 10898 }, { "epoch": 1.3864648263579697, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.26469898223877, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8559505939483643, "num_tokens": 416055648.0, "step": 10899 }, { "epoch": 1.3865920366365603, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.314804077148438, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8612237572669983, "num_tokens": 416087062.0, "step": 10900 }, { "epoch": 1.3867192469151508, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.278897285461426, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8760719299316406, "num_tokens": 416122559.0, "step": 10901 }, { "epoch": 1.3868464571937413, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.239274024963379, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8616383075714111, "num_tokens": 416155228.0, "step": 10902 }, { "epoch": 1.3869736674723319, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.317834854125977, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8553949594497681, "num_tokens": 416191190.0, "step": 10903 }, { "epoch": 1.3871008777509224, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.284865379333496, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8726414442062378, "num_tokens": 416228009.0, "step": 10904 }, { "epoch": 1.387228088029513, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.304265975952148, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8642828464508057, "num_tokens": 416263657.0, "step": 10905 }, { "epoch": 1.3873552983081032, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.236471176147461, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8628012537956238, "num_tokens": 416305187.0, "step": 10906 }, { "epoch": 1.3874825085866938, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.296591758728027, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8664842247962952, "num_tokens": 416342578.0, "step": 10907 }, { "epoch": 1.3876097188652843, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.252237319946289, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8668319582939148, "num_tokens": 416378969.0, "step": 10908 }, { "epoch": 1.3877369291438748, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.301899909973145, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8669651746749878, "num_tokens": 416412958.0, "step": 10909 }, { "epoch": 1.3878641394224653, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.24395751953125, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8499133586883545, "num_tokens": 416445643.0, "step": 10910 }, { "epoch": 1.3879913497010559, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.328226089477539, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8553254008293152, "num_tokens": 416484891.0, "step": 10911 }, { "epoch": 1.3881185599796464, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.24905014038086, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8550990223884583, "num_tokens": 416520504.0, "step": 10912 }, { "epoch": 1.388245770258237, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.227627754211426, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8690729737281799, "num_tokens": 416561877.0, "step": 10913 }, { "epoch": 1.3883729805368275, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.323527336120605, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8699833154678345, "num_tokens": 416597676.0, "step": 10914 }, { "epoch": 1.3885001908154178, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.240199089050293, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8747735023498535, "num_tokens": 416630557.0, "step": 10915 }, { "epoch": 1.3886274010940083, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.301359176635742, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8706948757171631, "num_tokens": 416668604.0, "step": 10916 }, { "epoch": 1.3887546113725988, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.232565879821777, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8481488227844238, "num_tokens": 416710246.0, "step": 10917 }, { "epoch": 1.3888818216511893, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.306469917297363, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8681281805038452, "num_tokens": 416744962.0, "step": 10918 }, { "epoch": 1.3890090319297799, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.261902809143066, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8695460557937622, "num_tokens": 416779483.0, "step": 10919 }, { "epoch": 1.3891362422083704, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.220430374145508, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.880860447883606, "num_tokens": 416822135.0, "step": 10920 }, { "epoch": 1.389263452486961, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.248560905456543, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8667059540748596, "num_tokens": 416860383.0, "step": 10921 }, { "epoch": 1.3893906627655515, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.219635009765625, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8644998073577881, "num_tokens": 416896960.0, "step": 10922 }, { "epoch": 1.389517873044142, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.279012680053711, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8796437382698059, "num_tokens": 416931981.0, "step": 10923 }, { "epoch": 1.3896450833227325, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.228800773620605, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8768366575241089, "num_tokens": 416968179.0, "step": 10924 }, { "epoch": 1.389772293601323, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.254535675048828, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8733465075492859, "num_tokens": 417008879.0, "step": 10925 }, { "epoch": 1.3898995038799136, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.214862823486328, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8789153099060059, "num_tokens": 417042497.0, "step": 10926 }, { "epoch": 1.390026714158504, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.207500457763672, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8700056672096252, "num_tokens": 417084396.0, "step": 10927 }, { "epoch": 1.3901539244370946, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.22413444519043, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.881618320941925, "num_tokens": 417121014.0, "step": 10928 }, { "epoch": 1.3902811347156852, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.252327919006348, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8678481578826904, "num_tokens": 417160169.0, "step": 10929 }, { "epoch": 1.3904083449942755, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.298336029052734, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8527552485466003, "num_tokens": 417195024.0, "step": 10930 }, { "epoch": 1.390535555272866, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.248146057128906, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8747614026069641, "num_tokens": 417231344.0, "step": 10931 }, { "epoch": 1.3906627655514565, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.279820442199707, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8603399395942688, "num_tokens": 417270266.0, "step": 10932 }, { "epoch": 1.390789975830047, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.240226745605469, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8669790029525757, "num_tokens": 417308292.0, "step": 10933 }, { "epoch": 1.3909171861086376, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.311156272888184, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8659183979034424, "num_tokens": 417346358.0, "step": 10934 }, { "epoch": 1.391044396387228, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.229461669921875, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8691124320030212, "num_tokens": 417381924.0, "step": 10935 }, { "epoch": 1.3911716066658186, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.324637413024902, "learning_rate": 1e-06, "loss": 0.4865, "mean_token_accuracy": 0.8424376249313354, "num_tokens": 417417779.0, "step": 10936 }, { "epoch": 1.3912988169444092, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.279899597167969, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8616746068000793, "num_tokens": 417459076.0, "step": 10937 }, { "epoch": 1.3914260272229997, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.231778144836426, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8683675527572632, "num_tokens": 417499422.0, "step": 10938 }, { "epoch": 1.39155323750159, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.303150177001953, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8686783313751221, "num_tokens": 417541776.0, "step": 10939 }, { "epoch": 1.3916804477801805, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.269664764404297, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8564500212669373, "num_tokens": 417576469.0, "step": 10940 }, { "epoch": 1.391807658058771, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.329682350158691, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.868604302406311, "num_tokens": 417611726.0, "step": 10941 }, { "epoch": 1.3919348683373616, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.231291770935059, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8612114787101746, "num_tokens": 417650135.0, "step": 10942 }, { "epoch": 1.392062078615952, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.288299560546875, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.849862813949585, "num_tokens": 417690981.0, "step": 10943 }, { "epoch": 1.3921892888945426, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.217853546142578, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.855743408203125, "num_tokens": 417732152.0, "step": 10944 }, { "epoch": 1.3923164991731332, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.296558380126953, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8574463725090027, "num_tokens": 417771814.0, "step": 10945 }, { "epoch": 1.3924437094517237, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.239566802978516, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8582570552825928, "num_tokens": 417810663.0, "step": 10946 }, { "epoch": 1.3925709197303142, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.321353912353516, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.871955931186676, "num_tokens": 417845664.0, "step": 10947 }, { "epoch": 1.3926981300089047, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.204894065856934, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8601391315460205, "num_tokens": 417889617.0, "step": 10948 }, { "epoch": 1.3928253402874953, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.280744552612305, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8655126094818115, "num_tokens": 417930820.0, "step": 10949 }, { "epoch": 1.3929525505660858, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.273872375488281, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8686010837554932, "num_tokens": 417966799.0, "step": 10950 }, { "epoch": 1.3930797608446763, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.242006301879883, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8604536056518555, "num_tokens": 418009328.0, "step": 10951 }, { "epoch": 1.3932069711232669, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.369274139404297, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8700698614120483, "num_tokens": 418048548.0, "step": 10952 }, { "epoch": 1.3933341814018574, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.308087348937988, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8657892942428589, "num_tokens": 418094947.0, "step": 10953 }, { "epoch": 1.393461391680448, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.242684364318848, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8587554693222046, "num_tokens": 418137144.0, "step": 10954 }, { "epoch": 1.3935886019590382, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.371026039123535, "learning_rate": 1e-06, "loss": 0.3443, "mean_token_accuracy": 0.8893547058105469, "num_tokens": 418171449.0, "step": 10955 }, { "epoch": 1.3937158122376287, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.281468391418457, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8697198629379272, "num_tokens": 418209854.0, "step": 10956 }, { "epoch": 1.3938430225162193, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.331550598144531, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8746477365493774, "num_tokens": 418245121.0, "step": 10957 }, { "epoch": 1.3939702327948098, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.272838592529297, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8706119060516357, "num_tokens": 418284158.0, "step": 10958 }, { "epoch": 1.3940974430734003, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.302087783813477, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8745245337486267, "num_tokens": 418324091.0, "step": 10959 }, { "epoch": 1.3942246533519909, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.295994758605957, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8683425188064575, "num_tokens": 418364764.0, "step": 10960 }, { "epoch": 1.3943518636305814, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.244226455688477, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8648242950439453, "num_tokens": 418403974.0, "step": 10961 }, { "epoch": 1.394479073909172, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.2982177734375, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8585348129272461, "num_tokens": 418446341.0, "step": 10962 }, { "epoch": 1.3946062841877624, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.270232200622559, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8679265379905701, "num_tokens": 418490989.0, "step": 10963 }, { "epoch": 1.3947334944663528, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.269070625305176, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8619994521141052, "num_tokens": 418524633.0, "step": 10964 }, { "epoch": 1.3948607047449433, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.388144493103027, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8673473000526428, "num_tokens": 418562576.0, "step": 10965 }, { "epoch": 1.3949879150235338, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.234270095825195, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8671492338180542, "num_tokens": 418596537.0, "step": 10966 }, { "epoch": 1.3951151253021243, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.369437217712402, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8654872179031372, "num_tokens": 418625800.0, "step": 10967 }, { "epoch": 1.3952423355807149, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.248345375061035, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.854486346244812, "num_tokens": 418663818.0, "step": 10968 }, { "epoch": 1.3953695458593054, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.329352378845215, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8728903532028198, "num_tokens": 418700607.0, "step": 10969 }, { "epoch": 1.395496756137896, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.2980375289917, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8613401651382446, "num_tokens": 418739357.0, "step": 10970 }, { "epoch": 1.3956239664164865, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.311039924621582, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8739153146743774, "num_tokens": 418769752.0, "step": 10971 }, { "epoch": 1.395751176695077, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.224689483642578, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8489341735839844, "num_tokens": 418808196.0, "step": 10972 }, { "epoch": 1.3958783869736675, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.282580375671387, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8830239772796631, "num_tokens": 418843335.0, "step": 10973 }, { "epoch": 1.396005597252258, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.29761791229248, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8763155937194824, "num_tokens": 418876591.0, "step": 10974 }, { "epoch": 1.3961328075308486, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.290088653564453, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8551276922225952, "num_tokens": 418910482.0, "step": 10975 }, { "epoch": 1.396260017809439, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.306754112243652, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8590072393417358, "num_tokens": 418942248.0, "step": 10976 }, { "epoch": 1.3963872280880296, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.260161399841309, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8711749315261841, "num_tokens": 418977770.0, "step": 10977 }, { "epoch": 1.3965144383666201, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.37209701538086, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8607891798019409, "num_tokens": 419016464.0, "step": 10978 }, { "epoch": 1.3966416486452105, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.293620109558105, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8591457009315491, "num_tokens": 419053254.0, "step": 10979 }, { "epoch": 1.396768858923801, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.245782852172852, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8621957302093506, "num_tokens": 419088940.0, "step": 10980 }, { "epoch": 1.3968960692023915, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.258670806884766, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8672243356704712, "num_tokens": 419127897.0, "step": 10981 }, { "epoch": 1.397023279480982, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.283618927001953, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8697401285171509, "num_tokens": 419170921.0, "step": 10982 }, { "epoch": 1.3971504897595726, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.285015106201172, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8737162351608276, "num_tokens": 419206032.0, "step": 10983 }, { "epoch": 1.397277700038163, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.25613021850586, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8819407224655151, "num_tokens": 419244240.0, "step": 10984 }, { "epoch": 1.3974049103167536, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.302983283996582, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8753791451454163, "num_tokens": 419274902.0, "step": 10985 }, { "epoch": 1.3975321205953442, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.264212608337402, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8794662356376648, "num_tokens": 419312295.0, "step": 10986 }, { "epoch": 1.3976593308739347, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.29352855682373, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8507717847824097, "num_tokens": 419350999.0, "step": 10987 }, { "epoch": 1.397786541152525, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.329052925109863, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8681625127792358, "num_tokens": 419386739.0, "step": 10988 }, { "epoch": 1.3979137514311155, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.206833839416504, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8743135929107666, "num_tokens": 419430353.0, "step": 10989 }, { "epoch": 1.398040961709706, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.355281829833984, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8641847372055054, "num_tokens": 419462910.0, "step": 10990 }, { "epoch": 1.3981681719882966, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.31889533996582, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8775569796562195, "num_tokens": 419498727.0, "step": 10991 }, { "epoch": 1.398295382266887, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.284797668457031, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8594428896903992, "num_tokens": 419534496.0, "step": 10992 }, { "epoch": 1.3984225925454776, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.255587577819824, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8641950488090515, "num_tokens": 419575760.0, "step": 10993 }, { "epoch": 1.3985498028240682, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.295251846313477, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8751749992370605, "num_tokens": 419616822.0, "step": 10994 }, { "epoch": 1.3986770131026587, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.293469429016113, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8748191595077515, "num_tokens": 419650425.0, "step": 10995 }, { "epoch": 1.3988042233812492, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.270707130432129, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8752778172492981, "num_tokens": 419685501.0, "step": 10996 }, { "epoch": 1.3989314336598397, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.338031768798828, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8700436353683472, "num_tokens": 419724471.0, "step": 10997 }, { "epoch": 1.3990586439384303, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.289504051208496, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8760281205177307, "num_tokens": 419763683.0, "step": 10998 }, { "epoch": 1.3991858542170208, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.301101684570312, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.87175053358078, "num_tokens": 419801238.0, "step": 10999 }, { "epoch": 1.3993130644956113, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.294437408447266, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8654625415802002, "num_tokens": 419832787.0, "step": 11000 }, { "epoch": 1.3994402747742019, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.27228832244873, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8577360510826111, "num_tokens": 419864294.0, "step": 11001 }, { "epoch": 1.3995674850527924, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.276205062866211, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8702816963195801, "num_tokens": 419900493.0, "step": 11002 }, { "epoch": 1.399694695331383, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.236815452575684, "learning_rate": 1e-06, "loss": 0.3582, "mean_token_accuracy": 0.8864293098449707, "num_tokens": 419938917.0, "step": 11003 }, { "epoch": 1.3998219056099732, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.23547649383545, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8838157057762146, "num_tokens": 419978771.0, "step": 11004 }, { "epoch": 1.3999491158885637, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.282254219055176, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8667706847190857, "num_tokens": 420011836.0, "step": 11005 }, { "epoch": 1.4000763261671543, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.281717300415039, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8476007580757141, "num_tokens": 420051618.0, "step": 11006 }, { "epoch": 1.4002035364457448, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.30937385559082, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8552708029747009, "num_tokens": 420089119.0, "step": 11007 }, { "epoch": 1.4003307467243353, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.259675025939941, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8804060220718384, "num_tokens": 420123706.0, "step": 11008 }, { "epoch": 1.4004579570029259, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.28687858581543, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8650013208389282, "num_tokens": 420156723.0, "step": 11009 }, { "epoch": 1.4005851672815164, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.237552642822266, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8679401874542236, "num_tokens": 420198519.0, "step": 11010 }, { "epoch": 1.400712377560107, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.273993492126465, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8810857534408569, "num_tokens": 420231722.0, "step": 11011 }, { "epoch": 1.4008395878386974, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.215559959411621, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8709908723831177, "num_tokens": 420272059.0, "step": 11012 }, { "epoch": 1.4009667981172877, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.246143341064453, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8742119073867798, "num_tokens": 420307695.0, "step": 11013 }, { "epoch": 1.4010940083958783, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.332148551940918, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8776795864105225, "num_tokens": 420345989.0, "step": 11014 }, { "epoch": 1.4012212186744688, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.255558967590332, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8747059106826782, "num_tokens": 420387194.0, "step": 11015 }, { "epoch": 1.4013484289530593, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.287665367126465, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.883447527885437, "num_tokens": 420427608.0, "step": 11016 }, { "epoch": 1.4014756392316499, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.256356239318848, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8743899464607239, "num_tokens": 420461363.0, "step": 11017 }, { "epoch": 1.4016028495102404, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.32048511505127, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8510664701461792, "num_tokens": 420503577.0, "step": 11018 }, { "epoch": 1.401730059788831, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.286781311035156, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8790924549102783, "num_tokens": 420538852.0, "step": 11019 }, { "epoch": 1.4018572700674214, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.319549560546875, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8659045696258545, "num_tokens": 420577495.0, "step": 11020 }, { "epoch": 1.401984480346012, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.270953178405762, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8681964874267578, "num_tokens": 420616123.0, "step": 11021 }, { "epoch": 1.4021116906246025, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.288052558898926, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8787489533424377, "num_tokens": 420653904.0, "step": 11022 }, { "epoch": 1.402238900903193, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.283341407775879, "learning_rate": 1e-06, "loss": 0.3456, "mean_token_accuracy": 0.8921200037002563, "num_tokens": 420693539.0, "step": 11023 }, { "epoch": 1.4023661111817836, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.271379470825195, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8717445135116577, "num_tokens": 420729046.0, "step": 11024 }, { "epoch": 1.402493321460374, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.389957427978516, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8721202611923218, "num_tokens": 420766143.0, "step": 11025 }, { "epoch": 1.4026205317389646, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.211139678955078, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8656143546104431, "num_tokens": 420806367.0, "step": 11026 }, { "epoch": 1.4027477420175551, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.290055274963379, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.874866247177124, "num_tokens": 420845081.0, "step": 11027 }, { "epoch": 1.4028749522961454, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.247535705566406, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8740572333335876, "num_tokens": 420885833.0, "step": 11028 }, { "epoch": 1.403002162574736, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.335458755493164, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.874686598777771, "num_tokens": 420916313.0, "step": 11029 }, { "epoch": 1.4031293728533265, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.258912086486816, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.851094663143158, "num_tokens": 420957011.0, "step": 11030 }, { "epoch": 1.403256583131917, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.269718170166016, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8730669617652893, "num_tokens": 420994815.0, "step": 11031 }, { "epoch": 1.4033837934105076, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.329048156738281, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8760991096496582, "num_tokens": 421033977.0, "step": 11032 }, { "epoch": 1.403511003689098, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.294577598571777, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8732532262802124, "num_tokens": 421068871.0, "step": 11033 }, { "epoch": 1.4036382139676886, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.258959770202637, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8697468042373657, "num_tokens": 421105680.0, "step": 11034 }, { "epoch": 1.4037654242462791, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.344099044799805, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8812390565872192, "num_tokens": 421149519.0, "step": 11035 }, { "epoch": 1.4038926345248697, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.286852836608887, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8702033758163452, "num_tokens": 421190195.0, "step": 11036 }, { "epoch": 1.40401984480346, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.335624694824219, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8650324940681458, "num_tokens": 421218637.0, "step": 11037 }, { "epoch": 1.4041470550820505, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.36280632019043, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8739065527915955, "num_tokens": 421261300.0, "step": 11038 }, { "epoch": 1.404274265360641, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.30073070526123, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8656339645385742, "num_tokens": 421297250.0, "step": 11039 }, { "epoch": 1.4044014756392316, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.25754451751709, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8619945645332336, "num_tokens": 421334213.0, "step": 11040 }, { "epoch": 1.404528685917822, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.33666706085205, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8762779235839844, "num_tokens": 421368341.0, "step": 11041 }, { "epoch": 1.4046558961964126, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.250590324401855, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8573037981987, "num_tokens": 421408278.0, "step": 11042 }, { "epoch": 1.4047831064750032, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.304411888122559, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8611129522323608, "num_tokens": 421442559.0, "step": 11043 }, { "epoch": 1.4049103167535937, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.342106819152832, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8674975633621216, "num_tokens": 421474014.0, "step": 11044 }, { "epoch": 1.4050375270321842, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.301300048828125, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8441144824028015, "num_tokens": 421511193.0, "step": 11045 }, { "epoch": 1.4051647373107747, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.313541412353516, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.870034396648407, "num_tokens": 421547895.0, "step": 11046 }, { "epoch": 1.4052919475893653, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.330244064331055, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8691096305847168, "num_tokens": 421582883.0, "step": 11047 }, { "epoch": 1.4054191578679558, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.319733619689941, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8706594109535217, "num_tokens": 421619838.0, "step": 11048 }, { "epoch": 1.4055463681465463, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.328598022460938, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8783754110336304, "num_tokens": 421659451.0, "step": 11049 }, { "epoch": 1.4056735784251368, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.203326225280762, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8658996820449829, "num_tokens": 421704826.0, "step": 11050 }, { "epoch": 1.4058007887037274, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.366952896118164, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8649258613586426, "num_tokens": 421748953.0, "step": 11051 }, { "epoch": 1.405927998982318, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.307939529418945, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8520057797431946, "num_tokens": 421786614.0, "step": 11052 }, { "epoch": 1.4060552092609082, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.336177825927734, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8635955452919006, "num_tokens": 421817291.0, "step": 11053 }, { "epoch": 1.4061824195394987, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.24710464477539, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.866666316986084, "num_tokens": 421858200.0, "step": 11054 }, { "epoch": 1.4063096298180893, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.297390937805176, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8693584203720093, "num_tokens": 421903343.0, "step": 11055 }, { "epoch": 1.4064368400966798, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.294570922851562, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8851090669631958, "num_tokens": 421941426.0, "step": 11056 }, { "epoch": 1.4065640503752703, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.249221801757812, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8658861517906189, "num_tokens": 421983992.0, "step": 11057 }, { "epoch": 1.4066912606538609, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.275887489318848, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8771710395812988, "num_tokens": 422015066.0, "step": 11058 }, { "epoch": 1.4068184709324514, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.286394119262695, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.871868371963501, "num_tokens": 422052392.0, "step": 11059 }, { "epoch": 1.406945681211042, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.24621295928955, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8651829361915588, "num_tokens": 422091453.0, "step": 11060 }, { "epoch": 1.4070728914896324, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.30674934387207, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8532068729400635, "num_tokens": 422131019.0, "step": 11061 }, { "epoch": 1.4072001017682227, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.25672435760498, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8625218868255615, "num_tokens": 422172139.0, "step": 11062 }, { "epoch": 1.4073273120468133, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.260727882385254, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8750260472297668, "num_tokens": 422209708.0, "step": 11063 }, { "epoch": 1.4074545223254038, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.287272453308105, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8695685863494873, "num_tokens": 422245091.0, "step": 11064 }, { "epoch": 1.4075817326039943, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.314230918884277, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8692193627357483, "num_tokens": 422280054.0, "step": 11065 }, { "epoch": 1.4077089428825849, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.250404357910156, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8754047751426697, "num_tokens": 422310036.0, "step": 11066 }, { "epoch": 1.4078361531611754, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.284479141235352, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8532564043998718, "num_tokens": 422347914.0, "step": 11067 }, { "epoch": 1.407963363439766, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.345247268676758, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8702781200408936, "num_tokens": 422383207.0, "step": 11068 }, { "epoch": 1.4080905737183564, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.303083419799805, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8737273216247559, "num_tokens": 422414675.0, "step": 11069 }, { "epoch": 1.408217783996947, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.299703598022461, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8572131991386414, "num_tokens": 422449746.0, "step": 11070 }, { "epoch": 1.4083449942755375, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.318761825561523, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8622422218322754, "num_tokens": 422487789.0, "step": 11071 }, { "epoch": 1.408472204554128, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.3207368850708, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8743744492530823, "num_tokens": 422521750.0, "step": 11072 }, { "epoch": 1.4085994148327186, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.3282470703125, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8781659603118896, "num_tokens": 422562923.0, "step": 11073 }, { "epoch": 1.408726625111309, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.327669143676758, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8587416410446167, "num_tokens": 422606832.0, "step": 11074 }, { "epoch": 1.4088538353898996, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.274209022521973, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8715628981590271, "num_tokens": 422644465.0, "step": 11075 }, { "epoch": 1.4089810456684901, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.248577117919922, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8767484426498413, "num_tokens": 422681204.0, "step": 11076 }, { "epoch": 1.4091082559470804, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.32925033569336, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8590402603149414, "num_tokens": 422716325.0, "step": 11077 }, { "epoch": 1.409235466225671, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.27168083190918, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.858240008354187, "num_tokens": 422753644.0, "step": 11078 }, { "epoch": 1.4093626765042615, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.31814193725586, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8685623407363892, "num_tokens": 422790893.0, "step": 11079 }, { "epoch": 1.409489886782852, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.352044105529785, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8601127862930298, "num_tokens": 422829400.0, "step": 11080 }, { "epoch": 1.4096170970614426, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.31949520111084, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.877586305141449, "num_tokens": 422867386.0, "step": 11081 }, { "epoch": 1.409744307340033, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.311492919921875, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8716543912887573, "num_tokens": 422910736.0, "step": 11082 }, { "epoch": 1.4098715176186236, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.287453651428223, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8731086850166321, "num_tokens": 422950302.0, "step": 11083 }, { "epoch": 1.4099987278972141, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.341229438781738, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8672188520431519, "num_tokens": 422985892.0, "step": 11084 }, { "epoch": 1.4101259381758047, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.317930221557617, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8612009882926941, "num_tokens": 423020075.0, "step": 11085 }, { "epoch": 1.410253148454395, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.352983474731445, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.872286319732666, "num_tokens": 423058260.0, "step": 11086 }, { "epoch": 1.4103803587329855, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.25485897064209, "learning_rate": 1e-06, "loss": 0.3351, "mean_token_accuracy": 0.8917598724365234, "num_tokens": 423092158.0, "step": 11087 }, { "epoch": 1.410507569011576, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.341379165649414, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8663536906242371, "num_tokens": 423123999.0, "step": 11088 }, { "epoch": 1.4106347792901666, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.304476737976074, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8644368648529053, "num_tokens": 423167111.0, "step": 11089 }, { "epoch": 1.410761989568757, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.2810697555542, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8617140054702759, "num_tokens": 423202060.0, "step": 11090 }, { "epoch": 1.4108891998473476, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.30865478515625, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8603517413139343, "num_tokens": 423241200.0, "step": 11091 }, { "epoch": 1.4110164101259381, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.28226089477539, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8775737285614014, "num_tokens": 423284713.0, "step": 11092 }, { "epoch": 1.4111436204045287, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.302594184875488, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8704453706741333, "num_tokens": 423322524.0, "step": 11093 }, { "epoch": 1.4112708306831192, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.291243553161621, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.869708240032196, "num_tokens": 423360551.0, "step": 11094 }, { "epoch": 1.4113980409617097, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.233563423156738, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8792093396186829, "num_tokens": 423395227.0, "step": 11095 }, { "epoch": 1.4115252512403003, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.361824989318848, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8547296524047852, "num_tokens": 423434757.0, "step": 11096 }, { "epoch": 1.4116524615188908, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.29996395111084, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8858346939086914, "num_tokens": 423464599.0, "step": 11097 }, { "epoch": 1.4117796717974813, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.325237274169922, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8613792657852173, "num_tokens": 423504327.0, "step": 11098 }, { "epoch": 1.4119068820760718, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.318068504333496, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8505653142929077, "num_tokens": 423543114.0, "step": 11099 }, { "epoch": 1.4120340923546624, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.297547340393066, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8568375110626221, "num_tokens": 423581727.0, "step": 11100 }, { "epoch": 1.412161302633253, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.333089828491211, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8731792569160461, "num_tokens": 423617073.0, "step": 11101 }, { "epoch": 1.4122885129118432, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.21590518951416, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8726022839546204, "num_tokens": 423664922.0, "step": 11102 }, { "epoch": 1.4124157231904337, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.320551872253418, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8600400686264038, "num_tokens": 423706990.0, "step": 11103 }, { "epoch": 1.4125429334690243, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.279324531555176, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8534427881240845, "num_tokens": 423745945.0, "step": 11104 }, { "epoch": 1.4126701437476148, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.273882865905762, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8740667104721069, "num_tokens": 423780824.0, "step": 11105 }, { "epoch": 1.4127973540262053, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.313652992248535, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8817545175552368, "num_tokens": 423818640.0, "step": 11106 }, { "epoch": 1.4129245643047958, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.262296676635742, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8779237866401672, "num_tokens": 423856657.0, "step": 11107 }, { "epoch": 1.4130517745833864, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.327827453613281, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8801599740982056, "num_tokens": 423900288.0, "step": 11108 }, { "epoch": 1.413178984861977, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.319241523742676, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8746442198753357, "num_tokens": 423935929.0, "step": 11109 }, { "epoch": 1.4133061951405674, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.380792617797852, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8667177557945251, "num_tokens": 423970719.0, "step": 11110 }, { "epoch": 1.4134334054191577, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.33045768737793, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8703104257583618, "num_tokens": 424012338.0, "step": 11111 }, { "epoch": 1.4135606156977483, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.315848350524902, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.851909875869751, "num_tokens": 424050112.0, "step": 11112 }, { "epoch": 1.4136878259763388, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.379301071166992, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8696645498275757, "num_tokens": 424092688.0, "step": 11113 }, { "epoch": 1.4138150362549293, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.282553672790527, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8659496903419495, "num_tokens": 424131355.0, "step": 11114 }, { "epoch": 1.4139422465335199, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.324238777160645, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8588544726371765, "num_tokens": 424177768.0, "step": 11115 }, { "epoch": 1.4140694568121104, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.323546409606934, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8819049000740051, "num_tokens": 424215834.0, "step": 11116 }, { "epoch": 1.414196667090701, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.224809646606445, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.872706949710846, "num_tokens": 424260002.0, "step": 11117 }, { "epoch": 1.4143238773692914, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.324122428894043, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8670852184295654, "num_tokens": 424299343.0, "step": 11118 }, { "epoch": 1.414451087647882, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.213772773742676, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8633641004562378, "num_tokens": 424340312.0, "step": 11119 }, { "epoch": 1.4145782979264725, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.287773132324219, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.876826286315918, "num_tokens": 424376658.0, "step": 11120 }, { "epoch": 1.414705508205063, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.296064376831055, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8791539072990417, "num_tokens": 424415499.0, "step": 11121 }, { "epoch": 1.4148327184836536, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.340741157531738, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8704584836959839, "num_tokens": 424453401.0, "step": 11122 }, { "epoch": 1.414959928762244, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.257408142089844, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8736605644226074, "num_tokens": 424490654.0, "step": 11123 }, { "epoch": 1.4150871390408346, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.327540397644043, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8778771162033081, "num_tokens": 424534469.0, "step": 11124 }, { "epoch": 1.4152143493194251, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.330416679382324, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8600156307220459, "num_tokens": 424570187.0, "step": 11125 }, { "epoch": 1.4153415595980154, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.287379264831543, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.849502682685852, "num_tokens": 424607898.0, "step": 11126 }, { "epoch": 1.415468769876606, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.331127166748047, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8717403411865234, "num_tokens": 424652156.0, "step": 11127 }, { "epoch": 1.4155959801551965, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.322622299194336, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.859738826751709, "num_tokens": 424688735.0, "step": 11128 }, { "epoch": 1.415723190433787, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.331472396850586, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8700960874557495, "num_tokens": 424732859.0, "step": 11129 }, { "epoch": 1.4158504007123776, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.30233097076416, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8619136810302734, "num_tokens": 424769381.0, "step": 11130 }, { "epoch": 1.415977610990968, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.339512825012207, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8706409931182861, "num_tokens": 424802362.0, "step": 11131 }, { "epoch": 1.4161048212695586, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.289250373840332, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8663411140441895, "num_tokens": 424837897.0, "step": 11132 }, { "epoch": 1.4162320315481491, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.30311393737793, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.881328821182251, "num_tokens": 424878154.0, "step": 11133 }, { "epoch": 1.4163592418267397, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.32370376586914, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8759769201278687, "num_tokens": 424915063.0, "step": 11134 }, { "epoch": 1.41648645210533, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.26424503326416, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8680657744407654, "num_tokens": 424955159.0, "step": 11135 }, { "epoch": 1.4166136623839205, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.278621673583984, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8613473773002625, "num_tokens": 424992573.0, "step": 11136 }, { "epoch": 1.416740872662511, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.31725025177002, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8731108903884888, "num_tokens": 425030452.0, "step": 11137 }, { "epoch": 1.4168680829411016, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.279620170593262, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8831295371055603, "num_tokens": 425067118.0, "step": 11138 }, { "epoch": 1.416995293219692, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.29145622253418, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8607960939407349, "num_tokens": 425101684.0, "step": 11139 }, { "epoch": 1.4171225034982826, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.319718360900879, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8786643743515015, "num_tokens": 425143127.0, "step": 11140 }, { "epoch": 1.4172497137768731, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.291966438293457, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8696197271347046, "num_tokens": 425173912.0, "step": 11141 }, { "epoch": 1.4173769240554637, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.288028717041016, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8738881945610046, "num_tokens": 425210146.0, "step": 11142 }, { "epoch": 1.4175041343340542, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.280781745910645, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8819198608398438, "num_tokens": 425246082.0, "step": 11143 }, { "epoch": 1.4176313446126447, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.321920394897461, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8659805059432983, "num_tokens": 425287470.0, "step": 11144 }, { "epoch": 1.4177585548912353, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.279009819030762, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8659662008285522, "num_tokens": 425326553.0, "step": 11145 }, { "epoch": 1.4178857651698258, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.285810470581055, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8626993894577026, "num_tokens": 425363105.0, "step": 11146 }, { "epoch": 1.4180129754484163, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.30886173248291, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8637142777442932, "num_tokens": 425402513.0, "step": 11147 }, { "epoch": 1.4181401857270068, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.296064376831055, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8697652220726013, "num_tokens": 425440132.0, "step": 11148 }, { "epoch": 1.4182673960055974, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.288361549377441, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8681826591491699, "num_tokens": 425481738.0, "step": 11149 }, { "epoch": 1.418394606284188, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.263113975524902, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8674461245536804, "num_tokens": 425516916.0, "step": 11150 }, { "epoch": 1.4185218165627782, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.299893379211426, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8683556914329529, "num_tokens": 425555951.0, "step": 11151 }, { "epoch": 1.4186490268413687, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.322924613952637, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8692281246185303, "num_tokens": 425595302.0, "step": 11152 }, { "epoch": 1.4187762371199593, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.338662147521973, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8719683885574341, "num_tokens": 425623281.0, "step": 11153 }, { "epoch": 1.4189034473985498, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.269437789916992, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8777252435684204, "num_tokens": 425659124.0, "step": 11154 }, { "epoch": 1.4190306576771403, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.275652885437012, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8576321601867676, "num_tokens": 425695809.0, "step": 11155 }, { "epoch": 1.4191578679557308, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.332662582397461, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8606463670730591, "num_tokens": 425731890.0, "step": 11156 }, { "epoch": 1.4192850782343214, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.255818367004395, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8772899508476257, "num_tokens": 425770899.0, "step": 11157 }, { "epoch": 1.419412288512912, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.338833808898926, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8795356154441833, "num_tokens": 425807812.0, "step": 11158 }, { "epoch": 1.4195394987915024, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.299185752868652, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.864905595779419, "num_tokens": 425848997.0, "step": 11159 }, { "epoch": 1.4196667090700927, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.300289154052734, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8789518475532532, "num_tokens": 425887816.0, "step": 11160 }, { "epoch": 1.4197939193486833, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.32994270324707, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8608214259147644, "num_tokens": 425925488.0, "step": 11161 }, { "epoch": 1.4199211296272738, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.280342102050781, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8758838176727295, "num_tokens": 425965616.0, "step": 11162 }, { "epoch": 1.4200483399058643, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.309587478637695, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8621201515197754, "num_tokens": 426005308.0, "step": 11163 }, { "epoch": 1.4201755501844548, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.321381568908691, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8650646805763245, "num_tokens": 426042339.0, "step": 11164 }, { "epoch": 1.4203027604630454, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.328373908996582, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8691458702087402, "num_tokens": 426081925.0, "step": 11165 }, { "epoch": 1.420429970741636, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.317082405090332, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8633486032485962, "num_tokens": 426120918.0, "step": 11166 }, { "epoch": 1.4205571810202264, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.329179763793945, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8578110337257385, "num_tokens": 426158711.0, "step": 11167 }, { "epoch": 1.420684391298817, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.39189624786377, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8652349710464478, "num_tokens": 426197034.0, "step": 11168 }, { "epoch": 1.4208116015774075, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.318741798400879, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8621307611465454, "num_tokens": 426239390.0, "step": 11169 }, { "epoch": 1.420938811855998, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.333014488220215, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8592015504837036, "num_tokens": 426285488.0, "step": 11170 }, { "epoch": 1.4210660221345885, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.358240127563477, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8563109636306763, "num_tokens": 426323955.0, "step": 11171 }, { "epoch": 1.421193232413179, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.306739807128906, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8865886330604553, "num_tokens": 426359460.0, "step": 11172 }, { "epoch": 1.4213204426917696, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.343903541564941, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8613649606704712, "num_tokens": 426397104.0, "step": 11173 }, { "epoch": 1.4214476529703601, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.334756851196289, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.877358078956604, "num_tokens": 426441349.0, "step": 11174 }, { "epoch": 1.4215748632489504, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.387788772583008, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8771515488624573, "num_tokens": 426478410.0, "step": 11175 }, { "epoch": 1.421702073527541, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.347476959228516, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8670161962509155, "num_tokens": 426519874.0, "step": 11176 }, { "epoch": 1.4218292838061315, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.35743522644043, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8571367263793945, "num_tokens": 426559970.0, "step": 11177 }, { "epoch": 1.421956494084722, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.453383445739746, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8700714707374573, "num_tokens": 426593897.0, "step": 11178 }, { "epoch": 1.4220837043633126, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.36190128326416, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8686805963516235, "num_tokens": 426633712.0, "step": 11179 }, { "epoch": 1.422210914641903, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.392081260681152, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8840658664703369, "num_tokens": 426669085.0, "step": 11180 }, { "epoch": 1.4223381249204936, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.295998573303223, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8813722133636475, "num_tokens": 426709260.0, "step": 11181 }, { "epoch": 1.4224653351990841, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.393048286437988, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8584535717964172, "num_tokens": 426747747.0, "step": 11182 }, { "epoch": 1.4225925454776747, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.40627670288086, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8563438653945923, "num_tokens": 426788185.0, "step": 11183 }, { "epoch": 1.422719755756265, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.350946426391602, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8710559606552124, "num_tokens": 426828663.0, "step": 11184 }, { "epoch": 1.4228469660348555, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.389439582824707, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8593882322311401, "num_tokens": 426869205.0, "step": 11185 }, { "epoch": 1.422974176313446, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.337069511413574, "learning_rate": 1e-06, "loss": 0.3545, "mean_token_accuracy": 0.8839168548583984, "num_tokens": 426900561.0, "step": 11186 }, { "epoch": 1.4231013865920366, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.368606567382812, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8834623694419861, "num_tokens": 426940618.0, "step": 11187 }, { "epoch": 1.423228596870627, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.409605026245117, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8605577945709229, "num_tokens": 426975509.0, "step": 11188 }, { "epoch": 1.4233558071492176, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.333406448364258, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8778478503227234, "num_tokens": 427011647.0, "step": 11189 }, { "epoch": 1.4234830174278081, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.338409423828125, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8662036657333374, "num_tokens": 427046631.0, "step": 11190 }, { "epoch": 1.4236102277063987, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.339218139648438, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8659631013870239, "num_tokens": 427081364.0, "step": 11191 }, { "epoch": 1.4237374379849892, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.404388427734375, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8708938360214233, "num_tokens": 427118667.0, "step": 11192 }, { "epoch": 1.4238646482635797, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.321808815002441, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8740568161010742, "num_tokens": 427155017.0, "step": 11193 }, { "epoch": 1.4239918585421703, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.385231018066406, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8617824912071228, "num_tokens": 427191741.0, "step": 11194 }, { "epoch": 1.4241190688207608, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.365046501159668, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8725664615631104, "num_tokens": 427226210.0, "step": 11195 }, { "epoch": 1.4242462790993513, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.361672401428223, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8655601143836975, "num_tokens": 427261699.0, "step": 11196 }, { "epoch": 1.4243734893779418, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.409927368164062, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.858883261680603, "num_tokens": 427297158.0, "step": 11197 }, { "epoch": 1.4245006996565324, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.352762222290039, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8541340827941895, "num_tokens": 427338234.0, "step": 11198 }, { "epoch": 1.424627909935123, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.368522644042969, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8633788824081421, "num_tokens": 427376727.0, "step": 11199 }, { "epoch": 1.4247551202137132, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.398333549499512, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8807036280632019, "num_tokens": 427417095.0, "step": 11200 }, { "epoch": 1.4248823304923037, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.39809513092041, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8528627157211304, "num_tokens": 427453243.0, "step": 11201 }, { "epoch": 1.4250095407708943, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.33584213256836, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8716504573822021, "num_tokens": 427490961.0, "step": 11202 }, { "epoch": 1.4251367510494848, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.402599334716797, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8781604170799255, "num_tokens": 427521487.0, "step": 11203 }, { "epoch": 1.4252639613280753, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.377513885498047, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.865373969078064, "num_tokens": 427556607.0, "step": 11204 }, { "epoch": 1.4253911716066658, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.314943313598633, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8697338104248047, "num_tokens": 427598940.0, "step": 11205 }, { "epoch": 1.4255183818852564, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.435178756713867, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8583190441131592, "num_tokens": 427641337.0, "step": 11206 }, { "epoch": 1.425645592163847, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.375426292419434, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8798254728317261, "num_tokens": 427680235.0, "step": 11207 }, { "epoch": 1.4257728024424374, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.431140899658203, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.868497371673584, "num_tokens": 427719918.0, "step": 11208 }, { "epoch": 1.4259000127210277, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.377421379089355, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8716343641281128, "num_tokens": 427758042.0, "step": 11209 }, { "epoch": 1.4260272229996183, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.425837516784668, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8709352016448975, "num_tokens": 427794591.0, "step": 11210 }, { "epoch": 1.4261544332782088, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.409110069274902, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8773090839385986, "num_tokens": 427828190.0, "step": 11211 }, { "epoch": 1.4262816435567993, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.355401039123535, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8622813820838928, "num_tokens": 427870363.0, "step": 11212 }, { "epoch": 1.4264088538353898, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.404468536376953, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8832985758781433, "num_tokens": 427910037.0, "step": 11213 }, { "epoch": 1.4265360641139804, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.364472389221191, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8682793378829956, "num_tokens": 427945782.0, "step": 11214 }, { "epoch": 1.426663274392571, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.406524658203125, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8759896755218506, "num_tokens": 427984635.0, "step": 11215 }, { "epoch": 1.4267904846711614, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.335076332092285, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8841536641120911, "num_tokens": 428024289.0, "step": 11216 }, { "epoch": 1.426917694949752, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.318209648132324, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8608136177062988, "num_tokens": 428067119.0, "step": 11217 }, { "epoch": 1.4270449052283425, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.339546203613281, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8682325482368469, "num_tokens": 428108150.0, "step": 11218 }, { "epoch": 1.427172115506933, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.353864669799805, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8805228471755981, "num_tokens": 428150839.0, "step": 11219 }, { "epoch": 1.4272993257855235, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.333128929138184, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.857851505279541, "num_tokens": 428194325.0, "step": 11220 }, { "epoch": 1.427426536064114, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.372137069702148, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8682495355606079, "num_tokens": 428230879.0, "step": 11221 }, { "epoch": 1.4275537463427046, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.318412780761719, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8804789781570435, "num_tokens": 428265795.0, "step": 11222 }, { "epoch": 1.4276809566212951, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.373908042907715, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8758968114852905, "num_tokens": 428300413.0, "step": 11223 }, { "epoch": 1.4278081668998854, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.357168197631836, "learning_rate": 1e-06, "loss": 0.3443, "mean_token_accuracy": 0.8904274702072144, "num_tokens": 428339293.0, "step": 11224 }, { "epoch": 1.427935377178476, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.380821228027344, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8812063932418823, "num_tokens": 428374158.0, "step": 11225 }, { "epoch": 1.4280625874570665, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.359591484069824, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.871070146560669, "num_tokens": 428411715.0, "step": 11226 }, { "epoch": 1.428189797735657, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.332707405090332, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8585963249206543, "num_tokens": 428450000.0, "step": 11227 }, { "epoch": 1.4283170080142475, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.417755126953125, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8759900331497192, "num_tokens": 428484546.0, "step": 11228 }, { "epoch": 1.428444218292838, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.268477439880371, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8711937069892883, "num_tokens": 428527519.0, "step": 11229 }, { "epoch": 1.4285714285714286, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.502448081970215, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8657447099685669, "num_tokens": 428568744.0, "step": 11230 }, { "epoch": 1.4286986388500191, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.404377937316895, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8663401007652283, "num_tokens": 428613050.0, "step": 11231 }, { "epoch": 1.4288258491286097, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.350764274597168, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8669949769973755, "num_tokens": 428647334.0, "step": 11232 }, { "epoch": 1.4289530594072, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.454676628112793, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8764330148696899, "num_tokens": 428681015.0, "step": 11233 }, { "epoch": 1.4290802696857905, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.39613151550293, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8633307218551636, "num_tokens": 428718903.0, "step": 11234 }, { "epoch": 1.429207479964381, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.423548698425293, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8786839842796326, "num_tokens": 428756038.0, "step": 11235 }, { "epoch": 1.4293346902429716, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.44251537322998, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8860783576965332, "num_tokens": 428793908.0, "step": 11236 }, { "epoch": 1.429461900521562, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.34197998046875, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8628093600273132, "num_tokens": 428830924.0, "step": 11237 }, { "epoch": 1.4295891108001526, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.40832233428955, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8666270971298218, "num_tokens": 428869083.0, "step": 11238 }, { "epoch": 1.4297163210787431, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.425620079040527, "learning_rate": 1e-06, "loss": 0.4835, "mean_token_accuracy": 0.8447709083557129, "num_tokens": 428903970.0, "step": 11239 }, { "epoch": 1.4298435313573337, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.386375427246094, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8737736940383911, "num_tokens": 428940710.0, "step": 11240 }, { "epoch": 1.4299707416359242, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.292482376098633, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8725892305374146, "num_tokens": 428981160.0, "step": 11241 }, { "epoch": 1.4300979519145147, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.351028442382812, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8715052604675293, "num_tokens": 429022464.0, "step": 11242 }, { "epoch": 1.4302251621931052, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.381141662597656, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8707259893417358, "num_tokens": 429060964.0, "step": 11243 }, { "epoch": 1.4303523724716958, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.318145751953125, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8709635734558105, "num_tokens": 429101389.0, "step": 11244 }, { "epoch": 1.4304795827502863, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.331936836242676, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8665211200714111, "num_tokens": 429135547.0, "step": 11245 }, { "epoch": 1.4306067930288768, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.313384056091309, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8729978203773499, "num_tokens": 429177504.0, "step": 11246 }, { "epoch": 1.4307340033074674, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.360273361206055, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8741147518157959, "num_tokens": 429217923.0, "step": 11247 }, { "epoch": 1.430861213586058, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.34665584564209, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8703578114509583, "num_tokens": 429251514.0, "step": 11248 }, { "epoch": 1.4309884238646482, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.387665748596191, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.85435551404953, "num_tokens": 429291747.0, "step": 11249 }, { "epoch": 1.4311156341432387, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.293701171875, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8663671016693115, "num_tokens": 429333480.0, "step": 11250 }, { "epoch": 1.4312428444218293, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.376775741577148, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.881057620048523, "num_tokens": 429366482.0, "step": 11251 }, { "epoch": 1.4313700547004198, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.366796493530273, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8619676828384399, "num_tokens": 429400704.0, "step": 11252 }, { "epoch": 1.4314972649790103, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.331910133361816, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8583486080169678, "num_tokens": 429437463.0, "step": 11253 }, { "epoch": 1.4316244752576008, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.397866249084473, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8599284887313843, "num_tokens": 429474496.0, "step": 11254 }, { "epoch": 1.4317516855361914, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.367359161376953, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8560656905174255, "num_tokens": 429514641.0, "step": 11255 }, { "epoch": 1.431878895814782, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.375513076782227, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8693875074386597, "num_tokens": 429554054.0, "step": 11256 }, { "epoch": 1.4320061060933724, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.32810115814209, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8654688596725464, "num_tokens": 429594004.0, "step": 11257 }, { "epoch": 1.4321333163719627, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.356450080871582, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8767237067222595, "num_tokens": 429633019.0, "step": 11258 }, { "epoch": 1.4322605266505533, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.369095802307129, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8702267408370972, "num_tokens": 429666869.0, "step": 11259 }, { "epoch": 1.4323877369291438, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.373550415039062, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8781623840332031, "num_tokens": 429700049.0, "step": 11260 }, { "epoch": 1.4325149472077343, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.413375854492188, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8672285079956055, "num_tokens": 429737517.0, "step": 11261 }, { "epoch": 1.4326421574863248, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.385199546813965, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8816114664077759, "num_tokens": 429778697.0, "step": 11262 }, { "epoch": 1.4327693677649154, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.441627502441406, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.869460940361023, "num_tokens": 429811883.0, "step": 11263 }, { "epoch": 1.432896578043506, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.379545211791992, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8752734661102295, "num_tokens": 429851748.0, "step": 11264 }, { "epoch": 1.4330237883220964, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.365958213806152, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8593617677688599, "num_tokens": 429891033.0, "step": 11265 }, { "epoch": 1.433150998600687, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.416749000549316, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8716025352478027, "num_tokens": 429935980.0, "step": 11266 }, { "epoch": 1.4332782088792775, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.293264389038086, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8577296733856201, "num_tokens": 429970874.0, "step": 11267 }, { "epoch": 1.433405419157868, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.448979377746582, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8539028167724609, "num_tokens": 430017909.0, "step": 11268 }, { "epoch": 1.4335326294364585, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.39217472076416, "learning_rate": 1e-06, "loss": 0.5162, "mean_token_accuracy": 0.843671441078186, "num_tokens": 430050641.0, "step": 11269 }, { "epoch": 1.433659839715049, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.41338062286377, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8736317157745361, "num_tokens": 430082985.0, "step": 11270 }, { "epoch": 1.4337870499936396, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.286738395690918, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8697851896286011, "num_tokens": 430122340.0, "step": 11271 }, { "epoch": 1.4339142602722301, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.392169952392578, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.856339693069458, "num_tokens": 430162140.0, "step": 11272 }, { "epoch": 1.4340414705508204, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.37540054321289, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8792482614517212, "num_tokens": 430201672.0, "step": 11273 }, { "epoch": 1.434168680829411, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.31501579284668, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8592629432678223, "num_tokens": 430239760.0, "step": 11274 }, { "epoch": 1.4342958911080015, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.402889251708984, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8521151542663574, "num_tokens": 430280527.0, "step": 11275 }, { "epoch": 1.434423101386592, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.410594940185547, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8647581338882446, "num_tokens": 430312788.0, "step": 11276 }, { "epoch": 1.4345503116651825, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.394819259643555, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8764724135398865, "num_tokens": 430348044.0, "step": 11277 }, { "epoch": 1.434677521943773, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.353531837463379, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8837808966636658, "num_tokens": 430385676.0, "step": 11278 }, { "epoch": 1.4348047322223636, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.361741065979004, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8658125400543213, "num_tokens": 430416976.0, "step": 11279 }, { "epoch": 1.4349319425009541, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.439537048339844, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8808417916297913, "num_tokens": 430460002.0, "step": 11280 }, { "epoch": 1.4350591527795447, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.266459465026855, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8477315902709961, "num_tokens": 430503172.0, "step": 11281 }, { "epoch": 1.435186363058135, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.431036949157715, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8563936948776245, "num_tokens": 430541221.0, "step": 11282 }, { "epoch": 1.4353135733367255, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.365525245666504, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8619348406791687, "num_tokens": 430579527.0, "step": 11283 }, { "epoch": 1.435440783615316, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.433205604553223, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8712594509124756, "num_tokens": 430617263.0, "step": 11284 }, { "epoch": 1.4355679938939065, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.404990196228027, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8820949792861938, "num_tokens": 430649961.0, "step": 11285 }, { "epoch": 1.435695204172497, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.403637886047363, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8636585474014282, "num_tokens": 430682599.0, "step": 11286 }, { "epoch": 1.4358224144510876, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.405051231384277, "learning_rate": 1e-06, "loss": 0.4959, "mean_token_accuracy": 0.8414442539215088, "num_tokens": 430721014.0, "step": 11287 }, { "epoch": 1.4359496247296781, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.389162063598633, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8668620586395264, "num_tokens": 430757014.0, "step": 11288 }, { "epoch": 1.4360768350082687, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.375677108764648, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.869178056716919, "num_tokens": 430796169.0, "step": 11289 }, { "epoch": 1.4362040452868592, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.391318321228027, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8671261072158813, "num_tokens": 430832726.0, "step": 11290 }, { "epoch": 1.4363312555654497, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.418338775634766, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.856940507888794, "num_tokens": 430866343.0, "step": 11291 }, { "epoch": 1.4364584658440402, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.363065719604492, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8739851117134094, "num_tokens": 430901622.0, "step": 11292 }, { "epoch": 1.4365856761226308, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.429765701293945, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8817048072814941, "num_tokens": 430937499.0, "step": 11293 }, { "epoch": 1.4367128864012213, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.320655822753906, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8566159009933472, "num_tokens": 430974048.0, "step": 11294 }, { "epoch": 1.4368400966798118, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.3443603515625, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8771206140518188, "num_tokens": 431010927.0, "step": 11295 }, { "epoch": 1.4369673069584024, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.393882751464844, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8535025715827942, "num_tokens": 431051731.0, "step": 11296 }, { "epoch": 1.4370945172369929, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.39908218383789, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8840705752372742, "num_tokens": 431087876.0, "step": 11297 }, { "epoch": 1.4372217275155832, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.3847074508667, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8627826571464539, "num_tokens": 431126217.0, "step": 11298 }, { "epoch": 1.4373489377941737, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.303771018981934, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8692998886108398, "num_tokens": 431168049.0, "step": 11299 }, { "epoch": 1.4374761480727642, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.428167343139648, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8665114641189575, "num_tokens": 431211143.0, "step": 11300 }, { "epoch": 1.4376033583513548, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.36913776397705, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.876034140586853, "num_tokens": 431255747.0, "step": 11301 }, { "epoch": 1.4377305686299453, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.407071113586426, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8707209825515747, "num_tokens": 431296618.0, "step": 11302 }, { "epoch": 1.4378577789085358, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.396238327026367, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.869672417640686, "num_tokens": 431336493.0, "step": 11303 }, { "epoch": 1.4379849891871264, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.426094055175781, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8664806485176086, "num_tokens": 431376200.0, "step": 11304 }, { "epoch": 1.438112199465717, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.379800796508789, "learning_rate": 1e-06, "loss": 0.4783, "mean_token_accuracy": 0.8486206531524658, "num_tokens": 431414015.0, "step": 11305 }, { "epoch": 1.4382394097443074, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.412386894226074, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8641161918640137, "num_tokens": 431453498.0, "step": 11306 }, { "epoch": 1.4383666200228977, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.391593933105469, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8667556047439575, "num_tokens": 431491126.0, "step": 11307 }, { "epoch": 1.4384938303014883, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.391839027404785, "learning_rate": 1e-06, "loss": 0.3374, "mean_token_accuracy": 0.8911254405975342, "num_tokens": 431528549.0, "step": 11308 }, { "epoch": 1.4386210405800788, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.410902976989746, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8741436004638672, "num_tokens": 431569435.0, "step": 11309 }, { "epoch": 1.4387482508586693, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.377320289611816, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.873356282711029, "num_tokens": 431606125.0, "step": 11310 }, { "epoch": 1.4388754611372598, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.469369888305664, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8634313344955444, "num_tokens": 431643517.0, "step": 11311 }, { "epoch": 1.4390026714158504, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.380410194396973, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8772934675216675, "num_tokens": 431682058.0, "step": 11312 }, { "epoch": 1.439129881694441, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.436792373657227, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8801088333129883, "num_tokens": 431721241.0, "step": 11313 }, { "epoch": 1.4392570919730314, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.321250915527344, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8809453248977661, "num_tokens": 431757809.0, "step": 11314 }, { "epoch": 1.439384302251622, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.379433631896973, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8569071292877197, "num_tokens": 431795931.0, "step": 11315 }, { "epoch": 1.4395115125302125, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.47767448425293, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8781700134277344, "num_tokens": 431830517.0, "step": 11316 }, { "epoch": 1.439638722808803, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.383478164672852, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8714700937271118, "num_tokens": 431866047.0, "step": 11317 }, { "epoch": 1.4397659330873935, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.469664573669434, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8576545119285583, "num_tokens": 431901602.0, "step": 11318 }, { "epoch": 1.439893143365984, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.412548065185547, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8643898963928223, "num_tokens": 431934989.0, "step": 11319 }, { "epoch": 1.4400203536445746, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.476875305175781, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8726980686187744, "num_tokens": 431968664.0, "step": 11320 }, { "epoch": 1.4401475639231651, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.355169296264648, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8796283602714539, "num_tokens": 432008025.0, "step": 11321 }, { "epoch": 1.4402747742017554, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.402493476867676, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8733987808227539, "num_tokens": 432043526.0, "step": 11322 }, { "epoch": 1.440401984480346, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.421010971069336, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8698036670684814, "num_tokens": 432085175.0, "step": 11323 }, { "epoch": 1.4405291947589365, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.388456344604492, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8604755997657776, "num_tokens": 432122379.0, "step": 11324 }, { "epoch": 1.440656405037527, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.432717323303223, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8711593151092529, "num_tokens": 432157690.0, "step": 11325 }, { "epoch": 1.4407836153161175, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.43990421295166, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8770530819892883, "num_tokens": 432188885.0, "step": 11326 }, { "epoch": 1.440910825594708, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.335436820983887, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8689760565757751, "num_tokens": 432228804.0, "step": 11327 }, { "epoch": 1.4410380358732986, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.445392608642578, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8519315719604492, "num_tokens": 432264150.0, "step": 11328 }, { "epoch": 1.4411652461518891, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.31801986694336, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8621745705604553, "num_tokens": 432306759.0, "step": 11329 }, { "epoch": 1.4412924564304797, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.444628715515137, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8793624639511108, "num_tokens": 432346928.0, "step": 11330 }, { "epoch": 1.44141966670907, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.349709510803223, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8750892281532288, "num_tokens": 432386630.0, "step": 11331 }, { "epoch": 1.4415468769876605, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.370779037475586, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8558145761489868, "num_tokens": 432427331.0, "step": 11332 }, { "epoch": 1.441674087266251, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.374561309814453, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8615432977676392, "num_tokens": 432462901.0, "step": 11333 }, { "epoch": 1.4418012975448415, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.36081600189209, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.865608811378479, "num_tokens": 432494232.0, "step": 11334 }, { "epoch": 1.441928507823432, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.338385581970215, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8812652826309204, "num_tokens": 432534276.0, "step": 11335 }, { "epoch": 1.4420557181020226, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.339805603027344, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8610016703605652, "num_tokens": 432577337.0, "step": 11336 }, { "epoch": 1.4421829283806131, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.425773620605469, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.860244631767273, "num_tokens": 432614290.0, "step": 11337 }, { "epoch": 1.4423101386592037, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.376215934753418, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8696756958961487, "num_tokens": 432648558.0, "step": 11338 }, { "epoch": 1.4424373489377942, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.452524185180664, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8418881893157959, "num_tokens": 432684243.0, "step": 11339 }, { "epoch": 1.4425645592163847, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.427962303161621, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8675318956375122, "num_tokens": 432719355.0, "step": 11340 }, { "epoch": 1.4426917694949752, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.340174674987793, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8680438995361328, "num_tokens": 432758675.0, "step": 11341 }, { "epoch": 1.4428189797735658, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.391490936279297, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8657537698745728, "num_tokens": 432801677.0, "step": 11342 }, { "epoch": 1.4429461900521563, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.409173965454102, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8669770359992981, "num_tokens": 432836299.0, "step": 11343 }, { "epoch": 1.4430734003307468, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.417901992797852, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8668996691703796, "num_tokens": 432868838.0, "step": 11344 }, { "epoch": 1.4432006106093374, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.398192405700684, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8692399859428406, "num_tokens": 432904895.0, "step": 11345 }, { "epoch": 1.4433278208879279, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.437790870666504, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.86305832862854, "num_tokens": 432944785.0, "step": 11346 }, { "epoch": 1.4434550311665182, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.443374633789062, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.858384370803833, "num_tokens": 432975645.0, "step": 11347 }, { "epoch": 1.4435822414451087, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.332090377807617, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8632543683052063, "num_tokens": 433016887.0, "step": 11348 }, { "epoch": 1.4437094517236992, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.393745422363281, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.869712769985199, "num_tokens": 433053666.0, "step": 11349 }, { "epoch": 1.4438366620022898, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.324833869934082, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8634288311004639, "num_tokens": 433101131.0, "step": 11350 }, { "epoch": 1.4439638722808803, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.41026782989502, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8756575584411621, "num_tokens": 433132535.0, "step": 11351 }, { "epoch": 1.4440910825594708, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.39704418182373, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.853725016117096, "num_tokens": 433168229.0, "step": 11352 }, { "epoch": 1.4442182928380614, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.401188850402832, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.873504102230072, "num_tokens": 433206094.0, "step": 11353 }, { "epoch": 1.4443455031166519, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.445714950561523, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8558824062347412, "num_tokens": 433241038.0, "step": 11354 }, { "epoch": 1.4444727133952424, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.464262008666992, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8614957332611084, "num_tokens": 433274674.0, "step": 11355 }, { "epoch": 1.4445999236738327, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.354787826538086, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8713328838348389, "num_tokens": 433314511.0, "step": 11356 }, { "epoch": 1.4447271339524232, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.431166648864746, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8706028461456299, "num_tokens": 433349338.0, "step": 11357 }, { "epoch": 1.4448543442310138, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.449458122253418, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8788386583328247, "num_tokens": 433382609.0, "step": 11358 }, { "epoch": 1.4449815545096043, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.486011505126953, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8823214769363403, "num_tokens": 433411553.0, "step": 11359 }, { "epoch": 1.4451087647881948, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.343254089355469, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8660266399383545, "num_tokens": 433450571.0, "step": 11360 }, { "epoch": 1.4452359750667854, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.473134994506836, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8633155822753906, "num_tokens": 433486994.0, "step": 11361 }, { "epoch": 1.445363185345376, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.456750869750977, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8812956213951111, "num_tokens": 433521407.0, "step": 11362 }, { "epoch": 1.4454903956239664, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.398237228393555, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8797135353088379, "num_tokens": 433561338.0, "step": 11363 }, { "epoch": 1.445617605902557, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.512271881103516, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8758620023727417, "num_tokens": 433598892.0, "step": 11364 }, { "epoch": 1.4457448161811475, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.45775032043457, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8583652973175049, "num_tokens": 433641423.0, "step": 11365 }, { "epoch": 1.445872026459738, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.461302757263184, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8711347579956055, "num_tokens": 433677385.0, "step": 11366 }, { "epoch": 1.4459992367383285, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.402664184570312, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8667900562286377, "num_tokens": 433716901.0, "step": 11367 }, { "epoch": 1.446126447016919, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.47908878326416, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8640269041061401, "num_tokens": 433755099.0, "step": 11368 }, { "epoch": 1.4462536572955096, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.440054893493652, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8645898103713989, "num_tokens": 433794012.0, "step": 11369 }, { "epoch": 1.4463808675741001, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.400298118591309, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8718779683113098, "num_tokens": 433831965.0, "step": 11370 }, { "epoch": 1.4465080778526904, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.462800025939941, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8691748976707458, "num_tokens": 433870819.0, "step": 11371 }, { "epoch": 1.446635288131281, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.371930122375488, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8626700043678284, "num_tokens": 433909197.0, "step": 11372 }, { "epoch": 1.4467624984098715, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.470084190368652, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8743801116943359, "num_tokens": 433946026.0, "step": 11373 }, { "epoch": 1.446889708688462, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.382240295410156, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.8901885151863098, "num_tokens": 433986239.0, "step": 11374 }, { "epoch": 1.4470169189670525, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.405610084533691, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8734031915664673, "num_tokens": 434022557.0, "step": 11375 }, { "epoch": 1.447144129245643, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.469895362854004, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8720306158065796, "num_tokens": 434059258.0, "step": 11376 }, { "epoch": 1.4472713395242336, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.397153854370117, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8546689748764038, "num_tokens": 434094872.0, "step": 11377 }, { "epoch": 1.4473985498028241, "ewc_loss": 0.028076171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.41408920288086, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.845909595489502, "num_tokens": 434134842.0, "step": 11378 }, { "epoch": 1.4475257600814146, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.397000312805176, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8802869915962219, "num_tokens": 434173152.0, "step": 11379 }, { "epoch": 1.447652970360005, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.399999618530273, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8707476854324341, "num_tokens": 434208318.0, "step": 11380 }, { "epoch": 1.4477801806385955, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.4465913772583, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8659065961837769, "num_tokens": 434243693.0, "step": 11381 }, { "epoch": 1.447907390917186, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.403972625732422, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8571529388427734, "num_tokens": 434283172.0, "step": 11382 }, { "epoch": 1.4480346011957765, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.402462005615234, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8694784641265869, "num_tokens": 434318789.0, "step": 11383 }, { "epoch": 1.448161811474367, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.390422821044922, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8601487278938293, "num_tokens": 434358943.0, "step": 11384 }, { "epoch": 1.4482890217529576, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.45028305053711, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8697631359100342, "num_tokens": 434398264.0, "step": 11385 }, { "epoch": 1.4484162320315481, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.441521644592285, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8780611753463745, "num_tokens": 434433602.0, "step": 11386 }, { "epoch": 1.4485434423101387, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.319851875305176, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8618464469909668, "num_tokens": 434480001.0, "step": 11387 }, { "epoch": 1.4486706525887292, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.38749885559082, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8841202259063721, "num_tokens": 434519285.0, "step": 11388 }, { "epoch": 1.4487978628673197, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.420677185058594, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8676623106002808, "num_tokens": 434556118.0, "step": 11389 }, { "epoch": 1.4489250731459102, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.424843788146973, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8528929948806763, "num_tokens": 434595350.0, "step": 11390 }, { "epoch": 1.4490522834245008, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.380719184875488, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8662887811660767, "num_tokens": 434632988.0, "step": 11391 }, { "epoch": 1.4491794937030913, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.396286964416504, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8757208585739136, "num_tokens": 434673040.0, "step": 11392 }, { "epoch": 1.4493067039816818, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.385824203491211, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8842132091522217, "num_tokens": 434712201.0, "step": 11393 }, { "epoch": 1.4494339142602723, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.373998641967773, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8827376365661621, "num_tokens": 434756676.0, "step": 11394 }, { "epoch": 1.4495611245388629, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.369874954223633, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8724234104156494, "num_tokens": 434797890.0, "step": 11395 }, { "epoch": 1.4496883348174532, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.381491661071777, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.882648766040802, "num_tokens": 434836933.0, "step": 11396 }, { "epoch": 1.4498155450960437, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.388239860534668, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8817536234855652, "num_tokens": 434868098.0, "step": 11397 }, { "epoch": 1.4499427553746342, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.371322631835938, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8645684719085693, "num_tokens": 434907239.0, "step": 11398 }, { "epoch": 1.4500699656532248, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.386787414550781, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8745483160018921, "num_tokens": 434945672.0, "step": 11399 }, { "epoch": 1.4501971759318153, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.414255142211914, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.85853111743927, "num_tokens": 434981693.0, "step": 11400 }, { "epoch": 1.4503243862104058, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.361809730529785, "learning_rate": 1e-06, "loss": 0.334, "mean_token_accuracy": 0.8905174136161804, "num_tokens": 435012452.0, "step": 11401 }, { "epoch": 1.4504515964889964, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.350377082824707, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8620772361755371, "num_tokens": 435052711.0, "step": 11402 }, { "epoch": 1.4505788067675869, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.414571762084961, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8678494691848755, "num_tokens": 435092766.0, "step": 11403 }, { "epoch": 1.4507060170461774, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.346002578735352, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8544654846191406, "num_tokens": 435132093.0, "step": 11404 }, { "epoch": 1.4508332273247677, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.405128479003906, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8656140565872192, "num_tokens": 435169071.0, "step": 11405 }, { "epoch": 1.4509604376033582, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.336050987243652, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8792869448661804, "num_tokens": 435202665.0, "step": 11406 }, { "epoch": 1.4510876478819488, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.464621543884277, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8834051489830017, "num_tokens": 435234719.0, "step": 11407 }, { "epoch": 1.4512148581605393, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.366827011108398, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.883519172668457, "num_tokens": 435267090.0, "step": 11408 }, { "epoch": 1.4513420684391298, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.398653030395508, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8765954971313477, "num_tokens": 435307243.0, "step": 11409 }, { "epoch": 1.4514692787177204, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.349494934082031, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8765463829040527, "num_tokens": 435347507.0, "step": 11410 }, { "epoch": 1.4515964889963109, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.386404991149902, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.860222578048706, "num_tokens": 435383499.0, "step": 11411 }, { "epoch": 1.4517236992749014, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.379093170166016, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8687639236450195, "num_tokens": 435423500.0, "step": 11412 }, { "epoch": 1.451850909553492, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.361125946044922, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8731722831726074, "num_tokens": 435457075.0, "step": 11413 }, { "epoch": 1.4519781198320825, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.401360511779785, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8675671815872192, "num_tokens": 435498933.0, "step": 11414 }, { "epoch": 1.452105330110673, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.361856460571289, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8608605861663818, "num_tokens": 435535949.0, "step": 11415 }, { "epoch": 1.4522325403892635, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.453033447265625, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8624063730239868, "num_tokens": 435565252.0, "step": 11416 }, { "epoch": 1.452359750667854, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.340250968933105, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8501070737838745, "num_tokens": 435604663.0, "step": 11417 }, { "epoch": 1.4524869609464446, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.3925199508667, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8639522790908813, "num_tokens": 435640883.0, "step": 11418 }, { "epoch": 1.4526141712250351, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.397841453552246, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8717503547668457, "num_tokens": 435682787.0, "step": 11419 }, { "epoch": 1.4527413815036254, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.40444564819336, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8667792081832886, "num_tokens": 435719982.0, "step": 11420 }, { "epoch": 1.452868591782216, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.42926025390625, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8518089056015015, "num_tokens": 435755153.0, "step": 11421 }, { "epoch": 1.4529958020608065, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.453361511230469, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8778349161148071, "num_tokens": 435797513.0, "step": 11422 }, { "epoch": 1.453123012339397, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.373435020446777, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8692867755889893, "num_tokens": 435836835.0, "step": 11423 }, { "epoch": 1.4532502226179875, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.450461387634277, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8594739437103271, "num_tokens": 435875124.0, "step": 11424 }, { "epoch": 1.453377432896578, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.394792556762695, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8810075521469116, "num_tokens": 435914993.0, "step": 11425 }, { "epoch": 1.4535046431751686, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.388816833496094, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8764796257019043, "num_tokens": 435950916.0, "step": 11426 }, { "epoch": 1.4536318534537591, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.411627769470215, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.864749550819397, "num_tokens": 435991986.0, "step": 11427 }, { "epoch": 1.4537590637323496, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.513749122619629, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8772076368331909, "num_tokens": 436024603.0, "step": 11428 }, { "epoch": 1.45388627401094, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.39499568939209, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.864695131778717, "num_tokens": 436065540.0, "step": 11429 }, { "epoch": 1.4540134842895305, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.461569786071777, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8555387258529663, "num_tokens": 436104394.0, "step": 11430 }, { "epoch": 1.454140694568121, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.434728622436523, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.855822741985321, "num_tokens": 436139977.0, "step": 11431 }, { "epoch": 1.4542679048467115, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.376449584960938, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.884540319442749, "num_tokens": 436178047.0, "step": 11432 }, { "epoch": 1.454395115125302, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.47919750213623, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.867145836353302, "num_tokens": 436218331.0, "step": 11433 }, { "epoch": 1.4545223254038926, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.311700820922852, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8667993545532227, "num_tokens": 436255663.0, "step": 11434 }, { "epoch": 1.4546495356824831, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.427929878234863, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8685320019721985, "num_tokens": 436301328.0, "step": 11435 }, { "epoch": 1.4547767459610736, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.357842445373535, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8774775862693787, "num_tokens": 436336487.0, "step": 11436 }, { "epoch": 1.4549039562396642, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.403343200683594, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.870940089225769, "num_tokens": 436377629.0, "step": 11437 }, { "epoch": 1.4550311665182547, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.354560852050781, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8819450736045837, "num_tokens": 436421590.0, "step": 11438 }, { "epoch": 1.4551583767968452, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.46823787689209, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8619503378868103, "num_tokens": 436461172.0, "step": 11439 }, { "epoch": 1.4552855870754358, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.411309242248535, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8774339556694031, "num_tokens": 436497536.0, "step": 11440 }, { "epoch": 1.4554127973540263, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.395072937011719, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8655577301979065, "num_tokens": 436543906.0, "step": 11441 }, { "epoch": 1.4555400076326168, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.504090309143066, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8702881336212158, "num_tokens": 436579462.0, "step": 11442 }, { "epoch": 1.4556672179112073, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.463656425476074, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.861375093460083, "num_tokens": 436618840.0, "step": 11443 }, { "epoch": 1.4557944281897979, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.426474571228027, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8570135831832886, "num_tokens": 436654176.0, "step": 11444 }, { "epoch": 1.4559216384683882, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.4569091796875, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8802499771118164, "num_tokens": 436685507.0, "step": 11445 }, { "epoch": 1.4560488487469787, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.354239463806152, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8778444528579712, "num_tokens": 436729881.0, "step": 11446 }, { "epoch": 1.4561760590255692, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.403761863708496, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8677262663841248, "num_tokens": 436772355.0, "step": 11447 }, { "epoch": 1.4563032693041598, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.39423656463623, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8619171977043152, "num_tokens": 436808701.0, "step": 11448 }, { "epoch": 1.4564304795827503, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.40390396118164, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8780640363693237, "num_tokens": 436845441.0, "step": 11449 }, { "epoch": 1.4565576898613408, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.436816215515137, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8713291883468628, "num_tokens": 436885316.0, "step": 11450 }, { "epoch": 1.4566849001399313, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.417367935180664, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8634376525878906, "num_tokens": 436918315.0, "step": 11451 }, { "epoch": 1.4568121104185219, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.418539047241211, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8606942892074585, "num_tokens": 436956039.0, "step": 11452 }, { "epoch": 1.4569393206971124, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.42160415649414, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8542321920394897, "num_tokens": 436991274.0, "step": 11453 }, { "epoch": 1.4570665309757027, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.448713302612305, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.880029559135437, "num_tokens": 437029963.0, "step": 11454 }, { "epoch": 1.4571937412542932, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.423232078552246, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8620805740356445, "num_tokens": 437067030.0, "step": 11455 }, { "epoch": 1.4573209515328838, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.38164234161377, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8755178451538086, "num_tokens": 437103363.0, "step": 11456 }, { "epoch": 1.4574481618114743, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.48164176940918, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.874915599822998, "num_tokens": 437131574.0, "step": 11457 }, { "epoch": 1.4575753720900648, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.421064376831055, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8610023260116577, "num_tokens": 437169537.0, "step": 11458 }, { "epoch": 1.4577025823686554, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.423138618469238, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8735204339027405, "num_tokens": 437208025.0, "step": 11459 }, { "epoch": 1.4578297926472459, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.381741523742676, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8795931339263916, "num_tokens": 437246642.0, "step": 11460 }, { "epoch": 1.4579570029258364, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.427953720092773, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8655929565429688, "num_tokens": 437285274.0, "step": 11461 }, { "epoch": 1.458084213204427, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.458438873291016, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8858819007873535, "num_tokens": 437325865.0, "step": 11462 }, { "epoch": 1.4582114234830175, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.395777702331543, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8745337724685669, "num_tokens": 437364093.0, "step": 11463 }, { "epoch": 1.458338633761608, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.446264266967773, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8567105531692505, "num_tokens": 437402413.0, "step": 11464 }, { "epoch": 1.4584658440401985, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.452178001403809, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8548247814178467, "num_tokens": 437442583.0, "step": 11465 }, { "epoch": 1.458593054318789, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.437963485717773, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8684400320053101, "num_tokens": 437485682.0, "step": 11466 }, { "epoch": 1.4587202645973796, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.340665817260742, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8588966727256775, "num_tokens": 437524859.0, "step": 11467 }, { "epoch": 1.45884747487597, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.509561538696289, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8605032563209534, "num_tokens": 437563590.0, "step": 11468 }, { "epoch": 1.4589746851545604, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.350655555725098, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8844687938690186, "num_tokens": 437601429.0, "step": 11469 }, { "epoch": 1.459101895433151, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.515774726867676, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.856458306312561, "num_tokens": 437639144.0, "step": 11470 }, { "epoch": 1.4592291057117415, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.398719787597656, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8771002292633057, "num_tokens": 437680846.0, "step": 11471 }, { "epoch": 1.459356315990332, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.549301147460938, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8702346682548523, "num_tokens": 437722156.0, "step": 11472 }, { "epoch": 1.4594835262689225, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.458697319030762, "learning_rate": 1e-06, "loss": 0.4897, "mean_token_accuracy": 0.8451467156410217, "num_tokens": 437766905.0, "step": 11473 }, { "epoch": 1.459610736547513, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.402968406677246, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8684631586074829, "num_tokens": 437801425.0, "step": 11474 }, { "epoch": 1.4597379468261036, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.500370025634766, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8572663068771362, "num_tokens": 437838126.0, "step": 11475 }, { "epoch": 1.4598651571046941, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.390312194824219, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8633038997650146, "num_tokens": 437876597.0, "step": 11476 }, { "epoch": 1.4599923673832846, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.479471206665039, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8801617622375488, "num_tokens": 437911852.0, "step": 11477 }, { "epoch": 1.460119577661875, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.378769874572754, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8690532445907593, "num_tokens": 437953418.0, "step": 11478 }, { "epoch": 1.4602467879404655, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.377315521240234, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8743151426315308, "num_tokens": 437989561.0, "step": 11479 }, { "epoch": 1.460373998219056, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.445496559143066, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8769437074661255, "num_tokens": 438033464.0, "step": 11480 }, { "epoch": 1.4605012084976465, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.432901382446289, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8723238706588745, "num_tokens": 438068676.0, "step": 11481 }, { "epoch": 1.460628418776237, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.46316909790039, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8508516550064087, "num_tokens": 438107867.0, "step": 11482 }, { "epoch": 1.4607556290548276, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.468832015991211, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8856076002120972, "num_tokens": 438144256.0, "step": 11483 }, { "epoch": 1.4608828393334181, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.455732345581055, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8789049983024597, "num_tokens": 438183458.0, "step": 11484 }, { "epoch": 1.4610100496120086, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.45833969116211, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8562332987785339, "num_tokens": 438216025.0, "step": 11485 }, { "epoch": 1.4611372598905992, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.440729141235352, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8790013790130615, "num_tokens": 438258185.0, "step": 11486 }, { "epoch": 1.4612644701691897, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.433121681213379, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8652642369270325, "num_tokens": 438298189.0, "step": 11487 }, { "epoch": 1.4613916804477802, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.445143699645996, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8572465181350708, "num_tokens": 438336593.0, "step": 11488 }, { "epoch": 1.4615188907263708, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.40084171295166, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8757243156433105, "num_tokens": 438374110.0, "step": 11489 }, { "epoch": 1.4616461010049613, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.45875358581543, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.853096604347229, "num_tokens": 438413062.0, "step": 11490 }, { "epoch": 1.4617733112835518, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.444091796875, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8714492321014404, "num_tokens": 438452055.0, "step": 11491 }, { "epoch": 1.4619005215621423, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.46458911895752, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8715626001358032, "num_tokens": 438494224.0, "step": 11492 }, { "epoch": 1.4620277318407329, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.406977653503418, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8654899001121521, "num_tokens": 438529790.0, "step": 11493 }, { "epoch": 1.4621549421193232, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.403704643249512, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8722983598709106, "num_tokens": 438571371.0, "step": 11494 }, { "epoch": 1.4622821523979137, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.401860237121582, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8545718193054199, "num_tokens": 438612383.0, "step": 11495 }, { "epoch": 1.4624093626765042, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.434234619140625, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8593804836273193, "num_tokens": 438653443.0, "step": 11496 }, { "epoch": 1.4625365729550948, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.47209358215332, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8569220304489136, "num_tokens": 438689565.0, "step": 11497 }, { "epoch": 1.4626637832336853, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.383835792541504, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8702853918075562, "num_tokens": 438727753.0, "step": 11498 }, { "epoch": 1.4627909935122758, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.427152633666992, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8661270141601562, "num_tokens": 438764667.0, "step": 11499 }, { "epoch": 1.4629182037908663, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.346960067749023, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8743264675140381, "num_tokens": 438801739.0, "step": 11500 }, { "epoch": 1.4630454140694569, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.381953239440918, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8702702522277832, "num_tokens": 438840028.0, "step": 11501 }, { "epoch": 1.4631726243480474, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.435820579528809, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8663750886917114, "num_tokens": 438880632.0, "step": 11502 }, { "epoch": 1.4632998346266377, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.422161102294922, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8651872873306274, "num_tokens": 438923040.0, "step": 11503 }, { "epoch": 1.4634270449052282, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.382999420166016, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8690305948257446, "num_tokens": 438958597.0, "step": 11504 }, { "epoch": 1.4635542551838188, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.424542427062988, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8571109771728516, "num_tokens": 438997909.0, "step": 11505 }, { "epoch": 1.4636814654624093, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.423669815063477, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.877177894115448, "num_tokens": 439030084.0, "step": 11506 }, { "epoch": 1.4638086757409998, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.391448974609375, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8706127405166626, "num_tokens": 439067770.0, "step": 11507 }, { "epoch": 1.4639358860195903, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.488550186157227, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8708801865577698, "num_tokens": 439102476.0, "step": 11508 }, { "epoch": 1.4640630962981809, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.375935554504395, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8785727620124817, "num_tokens": 439138660.0, "step": 11509 }, { "epoch": 1.4641903065767714, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.410433769226074, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8765784502029419, "num_tokens": 439181690.0, "step": 11510 }, { "epoch": 1.464317516855362, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.406096458435059, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.875138521194458, "num_tokens": 439218825.0, "step": 11511 }, { "epoch": 1.4644447271339525, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.434508323669434, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8727090358734131, "num_tokens": 439253505.0, "step": 11512 }, { "epoch": 1.464571937412543, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.400843620300293, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.862835705280304, "num_tokens": 439292764.0, "step": 11513 }, { "epoch": 1.4646991476911335, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.441034317016602, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8783320784568787, "num_tokens": 439335269.0, "step": 11514 }, { "epoch": 1.464826357969724, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.4387845993042, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8703724145889282, "num_tokens": 439371977.0, "step": 11515 }, { "epoch": 1.4649535682483146, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.433343887329102, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.868045449256897, "num_tokens": 439412710.0, "step": 11516 }, { "epoch": 1.465080778526905, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.430112838745117, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.868194580078125, "num_tokens": 439449637.0, "step": 11517 }, { "epoch": 1.4652079888054954, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.44017505645752, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8632573485374451, "num_tokens": 439495143.0, "step": 11518 }, { "epoch": 1.465335199084086, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.367037773132324, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8573511242866516, "num_tokens": 439532721.0, "step": 11519 }, { "epoch": 1.4654624093626765, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.525960922241211, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8779546618461609, "num_tokens": 439567938.0, "step": 11520 }, { "epoch": 1.465589619641267, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.479001998901367, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8775191307067871, "num_tokens": 439595720.0, "step": 11521 }, { "epoch": 1.4657168299198575, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.443718910217285, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8696038722991943, "num_tokens": 439634091.0, "step": 11522 }, { "epoch": 1.465844040198448, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.412610054016113, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.858108401298523, "num_tokens": 439671978.0, "step": 11523 }, { "epoch": 1.4659712504770386, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.363851547241211, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8651436567306519, "num_tokens": 439707748.0, "step": 11524 }, { "epoch": 1.466098460755629, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.410712242126465, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8749693632125854, "num_tokens": 439744060.0, "step": 11525 }, { "epoch": 1.4662256710342196, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.38140869140625, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.860985517501831, "num_tokens": 439789277.0, "step": 11526 }, { "epoch": 1.46635288131281, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.358119010925293, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8856937885284424, "num_tokens": 439827524.0, "step": 11527 }, { "epoch": 1.4664800915914005, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.410476684570312, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8657800555229187, "num_tokens": 439865181.0, "step": 11528 }, { "epoch": 1.466607301869991, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.441149711608887, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8615398406982422, "num_tokens": 439900071.0, "step": 11529 }, { "epoch": 1.4667345121485815, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.396485328674316, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8681609630584717, "num_tokens": 439940839.0, "step": 11530 }, { "epoch": 1.466861722427172, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.470276832580566, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8657739162445068, "num_tokens": 439978908.0, "step": 11531 }, { "epoch": 1.4669889327057626, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.373099327087402, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8687801957130432, "num_tokens": 440013448.0, "step": 11532 }, { "epoch": 1.467116142984353, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.481101989746094, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.872011125087738, "num_tokens": 440048377.0, "step": 11533 }, { "epoch": 1.4672433532629436, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.445737838745117, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8635978102684021, "num_tokens": 440089806.0, "step": 11534 }, { "epoch": 1.4673705635415342, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.468856811523438, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8751413226127625, "num_tokens": 440124209.0, "step": 11535 }, { "epoch": 1.4674977738201247, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.42664909362793, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8695074915885925, "num_tokens": 440158349.0, "step": 11536 }, { "epoch": 1.4676249840987152, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.451188087463379, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8706946969032288, "num_tokens": 440195220.0, "step": 11537 }, { "epoch": 1.4677521943773058, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.457945823669434, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.868588924407959, "num_tokens": 440235158.0, "step": 11538 }, { "epoch": 1.4678794046558963, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.429914474487305, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8643262386322021, "num_tokens": 440271879.0, "step": 11539 }, { "epoch": 1.4680066149344868, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.414142608642578, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8675257563591003, "num_tokens": 440306454.0, "step": 11540 }, { "epoch": 1.4681338252130773, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.50212287902832, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8564141988754272, "num_tokens": 440341185.0, "step": 11541 }, { "epoch": 1.4682610354916679, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.43716812133789, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8677209615707397, "num_tokens": 440379168.0, "step": 11542 }, { "epoch": 1.4683882457702582, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.439327239990234, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8703161478042603, "num_tokens": 440417945.0, "step": 11543 }, { "epoch": 1.4685154560488487, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.491198539733887, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8541277050971985, "num_tokens": 440457864.0, "step": 11544 }, { "epoch": 1.4686426663274392, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.451380729675293, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8646909594535828, "num_tokens": 440491709.0, "step": 11545 }, { "epoch": 1.4687698766060298, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.407510757446289, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.874937891960144, "num_tokens": 440533570.0, "step": 11546 }, { "epoch": 1.4688970868846203, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.489250183105469, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8748403191566467, "num_tokens": 440568360.0, "step": 11547 }, { "epoch": 1.4690242971632108, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.42702579498291, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8783171772956848, "num_tokens": 440606848.0, "step": 11548 }, { "epoch": 1.4691515074418013, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.617159843444824, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8821706771850586, "num_tokens": 440638254.0, "step": 11549 }, { "epoch": 1.4692787177203919, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.400333404541016, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8749858140945435, "num_tokens": 440674122.0, "step": 11550 }, { "epoch": 1.4694059279989824, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.472715377807617, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8834229111671448, "num_tokens": 440718362.0, "step": 11551 }, { "epoch": 1.4695331382775727, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.494625091552734, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.876024603843689, "num_tokens": 440755971.0, "step": 11552 }, { "epoch": 1.4696603485561632, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.499327659606934, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8668438196182251, "num_tokens": 440794328.0, "step": 11553 }, { "epoch": 1.4697875588347538, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.419955253601074, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8556756973266602, "num_tokens": 440833383.0, "step": 11554 }, { "epoch": 1.4699147691133443, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.416481971740723, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8757195472717285, "num_tokens": 440875502.0, "step": 11555 }, { "epoch": 1.4700419793919348, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.524009704589844, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8717189431190491, "num_tokens": 440915259.0, "step": 11556 }, { "epoch": 1.4701691896705253, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.424232482910156, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8669110536575317, "num_tokens": 440951660.0, "step": 11557 }, { "epoch": 1.4702963999491159, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.484498023986816, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8686826229095459, "num_tokens": 440992533.0, "step": 11558 }, { "epoch": 1.4704236102277064, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.439702987670898, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.857401430606842, "num_tokens": 441027360.0, "step": 11559 }, { "epoch": 1.470550820506297, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.467366218566895, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8755180835723877, "num_tokens": 441071746.0, "step": 11560 }, { "epoch": 1.4706780307848875, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.460887908935547, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8678765296936035, "num_tokens": 441105378.0, "step": 11561 }, { "epoch": 1.470805241063478, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.430010795593262, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8691520094871521, "num_tokens": 441150000.0, "step": 11562 }, { "epoch": 1.4709324513420685, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.405109405517578, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8618907332420349, "num_tokens": 441188426.0, "step": 11563 }, { "epoch": 1.471059661620659, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.42011833190918, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8753543496131897, "num_tokens": 441226028.0, "step": 11564 }, { "epoch": 1.4711868718992496, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.49150276184082, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8695865869522095, "num_tokens": 441265788.0, "step": 11565 }, { "epoch": 1.47131408217784, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.486100196838379, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8784152269363403, "num_tokens": 441297280.0, "step": 11566 }, { "epoch": 1.4714412924564304, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.517904281616211, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8675504326820374, "num_tokens": 441331229.0, "step": 11567 }, { "epoch": 1.471568502735021, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.445479393005371, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8723664879798889, "num_tokens": 441371064.0, "step": 11568 }, { "epoch": 1.4716957130136115, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.437519073486328, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8510740399360657, "num_tokens": 441405706.0, "step": 11569 }, { "epoch": 1.471822923292202, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.445425033569336, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8606748580932617, "num_tokens": 441445966.0, "step": 11570 }, { "epoch": 1.4719501335707925, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.518004417419434, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8530552387237549, "num_tokens": 441478916.0, "step": 11571 }, { "epoch": 1.472077343849383, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.496190071105957, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8656949996948242, "num_tokens": 441519245.0, "step": 11572 }, { "epoch": 1.4722045541279736, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.439878463745117, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8611006140708923, "num_tokens": 441557485.0, "step": 11573 }, { "epoch": 1.472331764406564, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.483692169189453, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8792089223861694, "num_tokens": 441596042.0, "step": 11574 }, { "epoch": 1.4724589746851546, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.440451622009277, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8781700134277344, "num_tokens": 441634391.0, "step": 11575 }, { "epoch": 1.472586184963745, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.510363578796387, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8610631823539734, "num_tokens": 441672201.0, "step": 11576 }, { "epoch": 1.4727133952423355, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.390641212463379, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8669991493225098, "num_tokens": 441713350.0, "step": 11577 }, { "epoch": 1.472840605520926, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.464991569519043, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8671583533287048, "num_tokens": 441754677.0, "step": 11578 }, { "epoch": 1.4729678157995165, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.43525505065918, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.858829140663147, "num_tokens": 441797037.0, "step": 11579 }, { "epoch": 1.473095026078107, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.463898658752441, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8671006560325623, "num_tokens": 441831661.0, "step": 11580 }, { "epoch": 1.4732222363566976, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.460906982421875, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8723384737968445, "num_tokens": 441876933.0, "step": 11581 }, { "epoch": 1.473349446635288, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.420194625854492, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8905584812164307, "num_tokens": 441918015.0, "step": 11582 }, { "epoch": 1.4734766569138786, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.501670837402344, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8638886213302612, "num_tokens": 441956573.0, "step": 11583 }, { "epoch": 1.4736038671924692, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.417236328125, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8829605579376221, "num_tokens": 442002064.0, "step": 11584 }, { "epoch": 1.4737310774710597, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.460100173950195, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8758120536804199, "num_tokens": 442040122.0, "step": 11585 }, { "epoch": 1.4738582877496502, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.503913879394531, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8714830875396729, "num_tokens": 442079698.0, "step": 11586 }, { "epoch": 1.4739854980282407, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.45044994354248, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8818014860153198, "num_tokens": 442118252.0, "step": 11587 }, { "epoch": 1.4741127083068313, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.525798797607422, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8719099164009094, "num_tokens": 442154675.0, "step": 11588 }, { "epoch": 1.4742399185854218, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.376588821411133, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8757598400115967, "num_tokens": 442189449.0, "step": 11589 }, { "epoch": 1.4743671288640123, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.53632926940918, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8716849088668823, "num_tokens": 442222844.0, "step": 11590 }, { "epoch": 1.4744943391426026, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.474315643310547, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.876885175704956, "num_tokens": 442258599.0, "step": 11591 }, { "epoch": 1.4746215494211932, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.43902587890625, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8759943246841431, "num_tokens": 442294775.0, "step": 11592 }, { "epoch": 1.4747487596997837, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.431374549865723, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8800185918807983, "num_tokens": 442330313.0, "step": 11593 }, { "epoch": 1.4748759699783742, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.438097953796387, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8679364919662476, "num_tokens": 442364988.0, "step": 11594 }, { "epoch": 1.4750031802569648, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.415273666381836, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8712013959884644, "num_tokens": 442406927.0, "step": 11595 }, { "epoch": 1.4751303905355553, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.447861671447754, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.87235426902771, "num_tokens": 442442092.0, "step": 11596 }, { "epoch": 1.4752576008141458, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.421263694763184, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8758832812309265, "num_tokens": 442483013.0, "step": 11597 }, { "epoch": 1.4753848110927363, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.420714378356934, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8757901787757874, "num_tokens": 442525052.0, "step": 11598 }, { "epoch": 1.4755120213713269, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.411407470703125, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8612321615219116, "num_tokens": 442562543.0, "step": 11599 }, { "epoch": 1.4756392316499174, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.432659149169922, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8749009370803833, "num_tokens": 442604976.0, "step": 11600 }, { "epoch": 1.4757664419285077, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.436250686645508, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8812161087989807, "num_tokens": 442642271.0, "step": 11601 }, { "epoch": 1.4758936522070982, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.444832801818848, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.867672860622406, "num_tokens": 442680508.0, "step": 11602 }, { "epoch": 1.4760208624856888, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.44081974029541, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8585189580917358, "num_tokens": 442718391.0, "step": 11603 }, { "epoch": 1.4761480727642793, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.494840621948242, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8805371522903442, "num_tokens": 442757688.0, "step": 11604 }, { "epoch": 1.4762752830428698, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.449737548828125, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8820657730102539, "num_tokens": 442796704.0, "step": 11605 }, { "epoch": 1.4764024933214603, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.41169261932373, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8801417350769043, "num_tokens": 442838791.0, "step": 11606 }, { "epoch": 1.4765297036000509, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.501887321472168, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8700861930847168, "num_tokens": 442882368.0, "step": 11607 }, { "epoch": 1.4766569138786414, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.462471961975098, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8690253496170044, "num_tokens": 442923768.0, "step": 11608 }, { "epoch": 1.476784124157232, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.446990013122559, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8836736083030701, "num_tokens": 442955656.0, "step": 11609 }, { "epoch": 1.4769113344358225, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.45427417755127, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8599584102630615, "num_tokens": 442997323.0, "step": 11610 }, { "epoch": 1.477038544714413, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.444354057312012, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8560221195220947, "num_tokens": 443033454.0, "step": 11611 }, { "epoch": 1.4771657549930035, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.402563095092773, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8692429065704346, "num_tokens": 443077902.0, "step": 11612 }, { "epoch": 1.477292965271594, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.411084175109863, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8608042001724243, "num_tokens": 443116515.0, "step": 11613 }, { "epoch": 1.4774201755501846, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.412137031555176, "learning_rate": 1e-06, "loss": 0.3221, "mean_token_accuracy": 0.8950841426849365, "num_tokens": 443154124.0, "step": 11614 }, { "epoch": 1.477547385828775, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.436020851135254, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.878502368927002, "num_tokens": 443195852.0, "step": 11615 }, { "epoch": 1.4776745961073654, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.43440055847168, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8637747168540955, "num_tokens": 443229850.0, "step": 11616 }, { "epoch": 1.477801806385956, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.460416793823242, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.861312210559845, "num_tokens": 443274481.0, "step": 11617 }, { "epoch": 1.4779290166645465, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.397623062133789, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8757777214050293, "num_tokens": 443311737.0, "step": 11618 }, { "epoch": 1.478056226943137, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.457624435424805, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8621112704277039, "num_tokens": 443345310.0, "step": 11619 }, { "epoch": 1.4781834372217275, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.430103302001953, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8682587146759033, "num_tokens": 443392869.0, "step": 11620 }, { "epoch": 1.478310647500318, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.409276008605957, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8681949377059937, "num_tokens": 443432842.0, "step": 11621 }, { "epoch": 1.4784378577789086, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.456210136413574, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8718118071556091, "num_tokens": 443465954.0, "step": 11622 }, { "epoch": 1.478565068057499, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.379103660583496, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8605719804763794, "num_tokens": 443508506.0, "step": 11623 }, { "epoch": 1.4786922783360896, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.505393981933594, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8678606748580933, "num_tokens": 443543575.0, "step": 11624 }, { "epoch": 1.47881948861468, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.391749382019043, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8726767897605896, "num_tokens": 443578504.0, "step": 11625 }, { "epoch": 1.4789466988932705, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.462189674377441, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8825812935829163, "num_tokens": 443614647.0, "step": 11626 }, { "epoch": 1.479073909171861, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.455527305603027, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8657993078231812, "num_tokens": 443649244.0, "step": 11627 }, { "epoch": 1.4792011194504515, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.43167781829834, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8654772043228149, "num_tokens": 443688374.0, "step": 11628 }, { "epoch": 1.479328329729042, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.465252876281738, "learning_rate": 1e-06, "loss": 0.5017, "mean_token_accuracy": 0.8406123518943787, "num_tokens": 443730929.0, "step": 11629 }, { "epoch": 1.4794555400076326, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.480194091796875, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8655096292495728, "num_tokens": 443763067.0, "step": 11630 }, { "epoch": 1.479582750286223, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.532416343688965, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8579050302505493, "num_tokens": 443801273.0, "step": 11631 }, { "epoch": 1.4797099605648136, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.506864547729492, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8564711809158325, "num_tokens": 443840394.0, "step": 11632 }, { "epoch": 1.4798371708434042, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.477269172668457, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8653771281242371, "num_tokens": 443877486.0, "step": 11633 }, { "epoch": 1.4799643811219947, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.487580299377441, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8683462738990784, "num_tokens": 443913763.0, "step": 11634 }, { "epoch": 1.4800915914005852, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.489439964294434, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8771514892578125, "num_tokens": 443946840.0, "step": 11635 }, { "epoch": 1.4802188016791757, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.47365665435791, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8546888828277588, "num_tokens": 443992439.0, "step": 11636 }, { "epoch": 1.4803460119577663, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.555641174316406, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8793008327484131, "num_tokens": 444029348.0, "step": 11637 }, { "epoch": 1.4804732222363568, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.485734939575195, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8771535158157349, "num_tokens": 444068386.0, "step": 11638 }, { "epoch": 1.4806004325149473, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.492136001586914, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8582018613815308, "num_tokens": 444110307.0, "step": 11639 }, { "epoch": 1.4807276427935376, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.508831977844238, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8695244789123535, "num_tokens": 444153919.0, "step": 11640 }, { "epoch": 1.4808548530721282, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.463133811950684, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8708102107048035, "num_tokens": 444190575.0, "step": 11641 }, { "epoch": 1.4809820633507187, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.527573585510254, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8627379536628723, "num_tokens": 444226108.0, "step": 11642 }, { "epoch": 1.4811092736293092, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.447924613952637, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8657483458518982, "num_tokens": 444268162.0, "step": 11643 }, { "epoch": 1.4812364839078997, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.540169715881348, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8555644750595093, "num_tokens": 444309273.0, "step": 11644 }, { "epoch": 1.4813636941864903, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.478265762329102, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8732893466949463, "num_tokens": 444338697.0, "step": 11645 }, { "epoch": 1.4814909044650808, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.502657890319824, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8789936304092407, "num_tokens": 444377954.0, "step": 11646 }, { "epoch": 1.4816181147436713, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.436413764953613, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8706226348876953, "num_tokens": 444416996.0, "step": 11647 }, { "epoch": 1.4817453250222619, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.467022895812988, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8755842447280884, "num_tokens": 444455550.0, "step": 11648 }, { "epoch": 1.4818725353008524, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.520591735839844, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.86517733335495, "num_tokens": 444493979.0, "step": 11649 }, { "epoch": 1.4819997455794427, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.423818588256836, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8680187463760376, "num_tokens": 444540622.0, "step": 11650 }, { "epoch": 1.4821269558580332, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.500821113586426, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8663767576217651, "num_tokens": 444577772.0, "step": 11651 }, { "epoch": 1.4822541661366238, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.527436256408691, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.856132984161377, "num_tokens": 444615523.0, "step": 11652 }, { "epoch": 1.4823813764152143, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.387505531311035, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.868127703666687, "num_tokens": 444659785.0, "step": 11653 }, { "epoch": 1.4825085866938048, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.511609077453613, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8648010492324829, "num_tokens": 444700806.0, "step": 11654 }, { "epoch": 1.4826357969723953, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.503324508666992, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8701049089431763, "num_tokens": 444734320.0, "step": 11655 }, { "epoch": 1.4827630072509859, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.494277000427246, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8631647229194641, "num_tokens": 444770215.0, "step": 11656 }, { "epoch": 1.4828902175295764, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.492257118225098, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8675965070724487, "num_tokens": 444805432.0, "step": 11657 }, { "epoch": 1.483017427808167, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.526097297668457, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8685045838356018, "num_tokens": 444832398.0, "step": 11658 }, { "epoch": 1.4831446380867574, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.41100788116455, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8728619813919067, "num_tokens": 444876128.0, "step": 11659 }, { "epoch": 1.483271848365348, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.506157875061035, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.864410400390625, "num_tokens": 444915188.0, "step": 11660 }, { "epoch": 1.4833990586439385, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.49160385131836, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8735371828079224, "num_tokens": 444953528.0, "step": 11661 }, { "epoch": 1.483526268922529, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.490205764770508, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8627955317497253, "num_tokens": 444996469.0, "step": 11662 }, { "epoch": 1.4836534792011196, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.488490104675293, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8852783441543579, "num_tokens": 445034298.0, "step": 11663 }, { "epoch": 1.48378068947971, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.51383113861084, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8677732944488525, "num_tokens": 445079010.0, "step": 11664 }, { "epoch": 1.4839078997583004, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.494497299194336, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8826878666877747, "num_tokens": 445116009.0, "step": 11665 }, { "epoch": 1.484035110036891, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.431150436401367, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8812756538391113, "num_tokens": 445156313.0, "step": 11666 }, { "epoch": 1.4841623203154815, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.49836254119873, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8703752160072327, "num_tokens": 445187424.0, "step": 11667 }, { "epoch": 1.484289530594072, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.467772483825684, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8701637387275696, "num_tokens": 445233448.0, "step": 11668 }, { "epoch": 1.4844167408726625, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.430569648742676, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8801659345626831, "num_tokens": 445273663.0, "step": 11669 }, { "epoch": 1.484543951151253, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.47314739227295, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8752095699310303, "num_tokens": 445313790.0, "step": 11670 }, { "epoch": 1.4846711614298436, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.482711791992188, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.879496693611145, "num_tokens": 445351818.0, "step": 11671 }, { "epoch": 1.484798371708434, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.525688171386719, "learning_rate": 1e-06, "loss": 0.489, "mean_token_accuracy": 0.8446858525276184, "num_tokens": 445391763.0, "step": 11672 }, { "epoch": 1.4849255819870246, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.498971939086914, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8626679182052612, "num_tokens": 445426723.0, "step": 11673 }, { "epoch": 1.485052792265615, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.474944114685059, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8742599487304688, "num_tokens": 445465635.0, "step": 11674 }, { "epoch": 1.4851800025442055, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.515335083007812, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8693743944168091, "num_tokens": 445502819.0, "step": 11675 }, { "epoch": 1.485307212822796, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.445778846740723, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8786159157752991, "num_tokens": 445541861.0, "step": 11676 }, { "epoch": 1.4854344231013865, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.543243408203125, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8845412731170654, "num_tokens": 445578058.0, "step": 11677 }, { "epoch": 1.485561633379977, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.538928031921387, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8548697829246521, "num_tokens": 445612896.0, "step": 11678 }, { "epoch": 1.4856888436585676, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.470934867858887, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.862544059753418, "num_tokens": 445647849.0, "step": 11679 }, { "epoch": 1.485816053937158, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.56216812133789, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8547155857086182, "num_tokens": 445689199.0, "step": 11680 }, { "epoch": 1.4859432642157486, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.426698684692383, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8678960800170898, "num_tokens": 445727869.0, "step": 11681 }, { "epoch": 1.4860704744943392, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.46919059753418, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8718979954719543, "num_tokens": 445770105.0, "step": 11682 }, { "epoch": 1.4861976847729297, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.455135345458984, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8487344980239868, "num_tokens": 445812950.0, "step": 11683 }, { "epoch": 1.4863248950515202, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.505918502807617, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8674397468566895, "num_tokens": 445850811.0, "step": 11684 }, { "epoch": 1.4864521053301107, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.432226181030273, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8715896010398865, "num_tokens": 445891967.0, "step": 11685 }, { "epoch": 1.4865793156087013, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.545998573303223, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8707538843154907, "num_tokens": 445927936.0, "step": 11686 }, { "epoch": 1.4867065258872918, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.426881790161133, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8812459707260132, "num_tokens": 445962481.0, "step": 11687 }, { "epoch": 1.4868337361658823, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.480842590332031, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8822294473648071, "num_tokens": 446001445.0, "step": 11688 }, { "epoch": 1.4869609464444726, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.470754623413086, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8873530030250549, "num_tokens": 446037904.0, "step": 11689 }, { "epoch": 1.4870881567230632, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.480033874511719, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8775987029075623, "num_tokens": 446077583.0, "step": 11690 }, { "epoch": 1.4872153670016537, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.502633094787598, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8678390383720398, "num_tokens": 446110123.0, "step": 11691 }, { "epoch": 1.4873425772802442, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.487445831298828, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8666313290596008, "num_tokens": 446149832.0, "step": 11692 }, { "epoch": 1.4874697875588347, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.446595191955566, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.875780463218689, "num_tokens": 446184211.0, "step": 11693 }, { "epoch": 1.4875969978374253, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.480273246765137, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8636295199394226, "num_tokens": 446229620.0, "step": 11694 }, { "epoch": 1.4877242081160158, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.523850440979004, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8708645701408386, "num_tokens": 446267134.0, "step": 11695 }, { "epoch": 1.4878514183946063, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.499872207641602, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.857801079750061, "num_tokens": 446303001.0, "step": 11696 }, { "epoch": 1.4879786286731969, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.492918014526367, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8678594827651978, "num_tokens": 446334286.0, "step": 11697 }, { "epoch": 1.4881058389517874, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.511908531188965, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8861973285675049, "num_tokens": 446369547.0, "step": 11698 }, { "epoch": 1.4882330492303777, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.443281173706055, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8549314737319946, "num_tokens": 446408179.0, "step": 11699 }, { "epoch": 1.4883602595089682, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.514533042907715, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8713586926460266, "num_tokens": 446448533.0, "step": 11700 }, { "epoch": 1.4884874697875587, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.451218605041504, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8670730590820312, "num_tokens": 446489445.0, "step": 11701 }, { "epoch": 1.4886146800661493, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.477676391601562, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8832083940505981, "num_tokens": 446526530.0, "step": 11702 }, { "epoch": 1.4887418903447398, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.527505874633789, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8698868751525879, "num_tokens": 446558315.0, "step": 11703 }, { "epoch": 1.4888691006233303, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.447003364562988, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8719956874847412, "num_tokens": 446598548.0, "step": 11704 }, { "epoch": 1.4889963109019209, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.513440132141113, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8653604388237, "num_tokens": 446632111.0, "step": 11705 }, { "epoch": 1.4891235211805114, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.564894676208496, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8774758577346802, "num_tokens": 446669296.0, "step": 11706 }, { "epoch": 1.489250731459102, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.477387428283691, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8705058693885803, "num_tokens": 446707891.0, "step": 11707 }, { "epoch": 1.4893779417376924, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.46580982208252, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8590325117111206, "num_tokens": 446749455.0, "step": 11708 }, { "epoch": 1.489505152016283, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.481451988220215, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8812735080718994, "num_tokens": 446787001.0, "step": 11709 }, { "epoch": 1.4896323622948735, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.477005004882812, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8694716095924377, "num_tokens": 446827262.0, "step": 11710 }, { "epoch": 1.489759572573464, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.565628051757812, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8844213485717773, "num_tokens": 446860514.0, "step": 11711 }, { "epoch": 1.4898867828520546, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.42835521697998, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8612295985221863, "num_tokens": 446902605.0, "step": 11712 }, { "epoch": 1.490013993130645, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.46530532836914, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8757225275039673, "num_tokens": 446938638.0, "step": 11713 }, { "epoch": 1.4901412034092354, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.441442489624023, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8745339512825012, "num_tokens": 446979286.0, "step": 11714 }, { "epoch": 1.490268413687826, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.476598739624023, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8545334935188293, "num_tokens": 447017405.0, "step": 11715 }, { "epoch": 1.4903956239664164, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.473766326904297, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8709472417831421, "num_tokens": 447056733.0, "step": 11716 }, { "epoch": 1.490522834245007, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.402583122253418, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8651822209358215, "num_tokens": 447094493.0, "step": 11717 }, { "epoch": 1.4906500445235975, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.459084510803223, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8855323791503906, "num_tokens": 447134039.0, "step": 11718 }, { "epoch": 1.490777254802188, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.453657150268555, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8778761029243469, "num_tokens": 447168320.0, "step": 11719 }, { "epoch": 1.4909044650807786, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.46971607208252, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8540836572647095, "num_tokens": 447205640.0, "step": 11720 }, { "epoch": 1.491031675359369, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.485095977783203, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8826470375061035, "num_tokens": 447239831.0, "step": 11721 }, { "epoch": 1.4911588856379596, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.4944429397583, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8553172945976257, "num_tokens": 447281100.0, "step": 11722 }, { "epoch": 1.49128609591655, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.453954696655273, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8506556749343872, "num_tokens": 447316137.0, "step": 11723 }, { "epoch": 1.4914133061951405, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.518610000610352, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8620447516441345, "num_tokens": 447357922.0, "step": 11724 }, { "epoch": 1.491540516473731, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.45090389251709, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8666865825653076, "num_tokens": 447394905.0, "step": 11725 }, { "epoch": 1.4916677267523215, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.5054349899292, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8797126412391663, "num_tokens": 447435379.0, "step": 11726 }, { "epoch": 1.491794937030912, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.515923500061035, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8603739738464355, "num_tokens": 447473058.0, "step": 11727 }, { "epoch": 1.4919221473095026, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.436478614807129, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8789957761764526, "num_tokens": 447510966.0, "step": 11728 }, { "epoch": 1.492049357588093, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.450366973876953, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8727271556854248, "num_tokens": 447554747.0, "step": 11729 }, { "epoch": 1.4921765678666836, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.529674530029297, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8722633123397827, "num_tokens": 447594426.0, "step": 11730 }, { "epoch": 1.4923037781452742, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.468605995178223, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8621307611465454, "num_tokens": 447634786.0, "step": 11731 }, { "epoch": 1.4924309884238647, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.462721824645996, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8667913675308228, "num_tokens": 447668682.0, "step": 11732 }, { "epoch": 1.4925581987024552, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.532990455627441, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8615782260894775, "num_tokens": 447705566.0, "step": 11733 }, { "epoch": 1.4926854089810457, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.414613723754883, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8663541078567505, "num_tokens": 447746385.0, "step": 11734 }, { "epoch": 1.4928126192596363, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.463791847229004, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.873775839805603, "num_tokens": 447790483.0, "step": 11735 }, { "epoch": 1.4929398295382268, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.434311866760254, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8670730590820312, "num_tokens": 447826872.0, "step": 11736 }, { "epoch": 1.4930670398168173, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512733459472656, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8738651275634766, "num_tokens": 447864071.0, "step": 11737 }, { "epoch": 1.4931942500954076, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.408520698547363, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8524854779243469, "num_tokens": 447903320.0, "step": 11738 }, { "epoch": 1.4933214603739982, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.517128944396973, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8829980492591858, "num_tokens": 447940253.0, "step": 11739 }, { "epoch": 1.4934486706525887, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.414632797241211, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8745895624160767, "num_tokens": 447978277.0, "step": 11740 }, { "epoch": 1.4935758809311792, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.568166732788086, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8578663468360901, "num_tokens": 448018798.0, "step": 11741 }, { "epoch": 1.4937030912097697, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.38083267211914, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8588401079177856, "num_tokens": 448057469.0, "step": 11742 }, { "epoch": 1.4938303014883603, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.489795684814453, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8569766283035278, "num_tokens": 448089032.0, "step": 11743 }, { "epoch": 1.4939575117669508, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.44279670715332, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8644986152648926, "num_tokens": 448122438.0, "step": 11744 }, { "epoch": 1.4940847220455413, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.402094841003418, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8678576946258545, "num_tokens": 448166844.0, "step": 11745 }, { "epoch": 1.4942119323241319, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.46120548248291, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.859595775604248, "num_tokens": 448197766.0, "step": 11746 }, { "epoch": 1.4943391426027224, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.444828033447266, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8570380210876465, "num_tokens": 448240850.0, "step": 11747 }, { "epoch": 1.4944663528813127, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.425517082214355, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.878333568572998, "num_tokens": 448274187.0, "step": 11748 }, { "epoch": 1.4945935631599032, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.529263496398926, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8749838471412659, "num_tokens": 448311069.0, "step": 11749 }, { "epoch": 1.4947207734384937, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.47254467010498, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8691859245300293, "num_tokens": 448351536.0, "step": 11750 }, { "epoch": 1.4948479837170843, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.535683631896973, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8660018444061279, "num_tokens": 448385108.0, "step": 11751 }, { "epoch": 1.4949751939956748, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.382906913757324, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8803343772888184, "num_tokens": 448423272.0, "step": 11752 }, { "epoch": 1.4951024042742653, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.471498489379883, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.874692440032959, "num_tokens": 448462904.0, "step": 11753 }, { "epoch": 1.4952296145528559, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.441471099853516, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.866222620010376, "num_tokens": 448503257.0, "step": 11754 }, { "epoch": 1.4953568248314464, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.550549507141113, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8744085431098938, "num_tokens": 448535299.0, "step": 11755 }, { "epoch": 1.495484035110037, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.459811210632324, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8753179907798767, "num_tokens": 448568878.0, "step": 11756 }, { "epoch": 1.4956112453886274, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.42981243133545, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8873294591903687, "num_tokens": 448604959.0, "step": 11757 }, { "epoch": 1.495738455667218, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.520384788513184, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8809051513671875, "num_tokens": 448638487.0, "step": 11758 }, { "epoch": 1.4958656659458085, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.45478343963623, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8749192953109741, "num_tokens": 448677212.0, "step": 11759 }, { "epoch": 1.495992876224399, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.415857315063477, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8643441200256348, "num_tokens": 448719654.0, "step": 11760 }, { "epoch": 1.4961200865029896, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.510708808898926, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8726717233657837, "num_tokens": 448760783.0, "step": 11761 }, { "epoch": 1.49624729678158, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.432063102722168, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8582308888435364, "num_tokens": 448802843.0, "step": 11762 }, { "epoch": 1.4963745070601704, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.534038543701172, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8662060499191284, "num_tokens": 448838430.0, "step": 11763 }, { "epoch": 1.496501717338761, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.434406280517578, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8599368333816528, "num_tokens": 448880450.0, "step": 11764 }, { "epoch": 1.4966289276173514, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.40777587890625, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8712083101272583, "num_tokens": 448919352.0, "step": 11765 }, { "epoch": 1.496756137895942, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.530816078186035, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8449369072914124, "num_tokens": 448965402.0, "step": 11766 }, { "epoch": 1.4968833481745325, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.444748878479004, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8806304931640625, "num_tokens": 448996983.0, "step": 11767 }, { "epoch": 1.497010558453123, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.519694328308105, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8571673035621643, "num_tokens": 449036555.0, "step": 11768 }, { "epoch": 1.4971377687317136, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.395136833190918, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8808208703994751, "num_tokens": 449070545.0, "step": 11769 }, { "epoch": 1.497264979010304, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.581232070922852, "learning_rate": 1e-06, "loss": 0.4737, "mean_token_accuracy": 0.8533260822296143, "num_tokens": 449108859.0, "step": 11770 }, { "epoch": 1.4973921892888946, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.485601425170898, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8659520149230957, "num_tokens": 449145333.0, "step": 11771 }, { "epoch": 1.497519399567485, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.540678024291992, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8631322979927063, "num_tokens": 449181292.0, "step": 11772 }, { "epoch": 1.4976466098460754, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.471196174621582, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8726850748062134, "num_tokens": 449213940.0, "step": 11773 }, { "epoch": 1.497773820124666, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.442805290222168, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8798117637634277, "num_tokens": 449252487.0, "step": 11774 }, { "epoch": 1.4979010304032565, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.516524314880371, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.862358808517456, "num_tokens": 449287268.0, "step": 11775 }, { "epoch": 1.498028240681847, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.474047660827637, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8661074638366699, "num_tokens": 449326975.0, "step": 11776 }, { "epoch": 1.4981554509604376, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.438982009887695, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8685832023620605, "num_tokens": 449366061.0, "step": 11777 }, { "epoch": 1.498282661239028, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.42845630645752, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8769121766090393, "num_tokens": 449401763.0, "step": 11778 }, { "epoch": 1.4984098715176186, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.506543159484863, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8880109786987305, "num_tokens": 449436605.0, "step": 11779 }, { "epoch": 1.4985370817962091, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.417874336242676, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.864517092704773, "num_tokens": 449474523.0, "step": 11780 }, { "epoch": 1.4986642920747997, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.454366683959961, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8814194798469543, "num_tokens": 449513301.0, "step": 11781 }, { "epoch": 1.4987915023533902, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.462655067443848, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.865159273147583, "num_tokens": 449547826.0, "step": 11782 }, { "epoch": 1.4989187126319807, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.464191436767578, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8686255812644958, "num_tokens": 449584818.0, "step": 11783 }, { "epoch": 1.4990459229105713, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.494349479675293, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8511220216751099, "num_tokens": 449625368.0, "step": 11784 }, { "epoch": 1.4991731331891618, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.456984519958496, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8569210171699524, "num_tokens": 449666320.0, "step": 11785 }, { "epoch": 1.4993003434677523, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.43692398071289, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8604989647865295, "num_tokens": 449705993.0, "step": 11786 }, { "epoch": 1.4994275537463426, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.45534610748291, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8648126125335693, "num_tokens": 449747088.0, "step": 11787 }, { "epoch": 1.4995547640249332, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.451837539672852, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8708593845367432, "num_tokens": 449789840.0, "step": 11788 }, { "epoch": 1.4996819743035237, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.442998886108398, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8682948350906372, "num_tokens": 449830599.0, "step": 11789 }, { "epoch": 1.4998091845821142, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.505748748779297, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.873441755771637, "num_tokens": 449864476.0, "step": 11790 }, { "epoch": 1.4999363948607047, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.48085880279541, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8533679842948914, "num_tokens": 449903791.0, "step": 11791 }, { "epoch": 1.5000636051392953, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.43733024597168, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.869186520576477, "num_tokens": 449945478.0, "step": 11792 }, { "epoch": 1.5001908154178858, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.45242691040039, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8659805059432983, "num_tokens": 449986175.0, "step": 11793 }, { "epoch": 1.5003180256964763, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.495113372802734, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8659285306930542, "num_tokens": 450025537.0, "step": 11794 }, { "epoch": 1.5004452359750666, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512564659118652, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8543647527694702, "num_tokens": 450068269.0, "step": 11795 }, { "epoch": 1.5005724462536572, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.534282684326172, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8847029209136963, "num_tokens": 450106234.0, "step": 11796 }, { "epoch": 1.5006996565322477, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.475589752197266, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8788800239562988, "num_tokens": 450141916.0, "step": 11797 }, { "epoch": 1.5008268668108382, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.525506973266602, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8551291227340698, "num_tokens": 450175427.0, "step": 11798 }, { "epoch": 1.5009540770894287, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.523015022277832, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.871433436870575, "num_tokens": 450211104.0, "step": 11799 }, { "epoch": 1.5010812873680193, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.499342918395996, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8610119819641113, "num_tokens": 450244844.0, "step": 11800 }, { "epoch": 1.5012084976466098, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.449953079223633, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8682793378829956, "num_tokens": 450286094.0, "step": 11801 }, { "epoch": 1.5013357079252003, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.719046592712402, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.860482394695282, "num_tokens": 450324938.0, "step": 11802 }, { "epoch": 1.5014629182037909, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.421613693237305, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8684592247009277, "num_tokens": 450368572.0, "step": 11803 }, { "epoch": 1.5015901284823814, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.597208976745605, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8476189374923706, "num_tokens": 450410926.0, "step": 11804 }, { "epoch": 1.501717338760972, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.51069450378418, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.872674286365509, "num_tokens": 450450206.0, "step": 11805 }, { "epoch": 1.5018445490395624, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.596722602844238, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.849286675453186, "num_tokens": 450488998.0, "step": 11806 }, { "epoch": 1.501971759318153, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.490619659423828, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.879459023475647, "num_tokens": 450530476.0, "step": 11807 }, { "epoch": 1.5020989695967435, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.475849151611328, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8760795593261719, "num_tokens": 450560655.0, "step": 11808 }, { "epoch": 1.502226179875334, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.502986907958984, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8719865679740906, "num_tokens": 450601816.0, "step": 11809 }, { "epoch": 1.5023533901539246, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.535174369812012, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8728322982788086, "num_tokens": 450634182.0, "step": 11810 }, { "epoch": 1.502480600432515, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.481133460998535, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8645533323287964, "num_tokens": 450672705.0, "step": 11811 }, { "epoch": 1.5026078107111056, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.533734321594238, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.855889081954956, "num_tokens": 450705636.0, "step": 11812 }, { "epoch": 1.502735020989696, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.487772941589355, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8597791194915771, "num_tokens": 450736165.0, "step": 11813 }, { "epoch": 1.5028622312682864, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.56929874420166, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8697911500930786, "num_tokens": 450776939.0, "step": 11814 }, { "epoch": 1.502989441546877, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.474773406982422, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8734704256057739, "num_tokens": 450811994.0, "step": 11815 }, { "epoch": 1.5031166518254675, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.542427062988281, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8699266314506531, "num_tokens": 450853688.0, "step": 11816 }, { "epoch": 1.503243862104058, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.502419471740723, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8692665100097656, "num_tokens": 450892100.0, "step": 11817 }, { "epoch": 1.5033710723826486, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.491903305053711, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8854166269302368, "num_tokens": 450936383.0, "step": 11818 }, { "epoch": 1.503498282661239, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.59658432006836, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8764712810516357, "num_tokens": 450965525.0, "step": 11819 }, { "epoch": 1.5036254929398294, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.491536140441895, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8567238450050354, "num_tokens": 451005874.0, "step": 11820 }, { "epoch": 1.50375270321842, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.46671199798584, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8491849303245544, "num_tokens": 451040668.0, "step": 11821 }, { "epoch": 1.5038799134970104, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.49539566040039, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8660968542098999, "num_tokens": 451082747.0, "step": 11822 }, { "epoch": 1.504007123775601, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.519458770751953, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8748162984848022, "num_tokens": 451120344.0, "step": 11823 }, { "epoch": 1.5041343340541915, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.501496315002441, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8624163269996643, "num_tokens": 451156564.0, "step": 11824 }, { "epoch": 1.504261544332782, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.413736343383789, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8637754321098328, "num_tokens": 451197252.0, "step": 11825 }, { "epoch": 1.5043887546113726, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.549595832824707, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8575817346572876, "num_tokens": 451236150.0, "step": 11826 }, { "epoch": 1.504515964889963, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.516022682189941, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8652846813201904, "num_tokens": 451275751.0, "step": 11827 }, { "epoch": 1.5046431751685536, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.493093490600586, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8887550830841064, "num_tokens": 451316778.0, "step": 11828 }, { "epoch": 1.5047703854471441, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.466547012329102, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8675045371055603, "num_tokens": 451359921.0, "step": 11829 }, { "epoch": 1.5048975957257347, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.45731258392334, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8758366107940674, "num_tokens": 451395616.0, "step": 11830 }, { "epoch": 1.5050248060043252, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.518128395080566, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8679736852645874, "num_tokens": 451427766.0, "step": 11831 }, { "epoch": 1.5051520162829157, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.581644058227539, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8774979114532471, "num_tokens": 451463073.0, "step": 11832 }, { "epoch": 1.5052792265615063, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.458475112915039, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8872842788696289, "num_tokens": 451503039.0, "step": 11833 }, { "epoch": 1.5054064368400968, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.622458457946777, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8515768051147461, "num_tokens": 451541599.0, "step": 11834 }, { "epoch": 1.5055336471186873, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.452065467834473, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8546539545059204, "num_tokens": 451579278.0, "step": 11835 }, { "epoch": 1.5056608573972778, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.55802059173584, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8719152212142944, "num_tokens": 451618591.0, "step": 11836 }, { "epoch": 1.5057880676758684, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 13.010473251342773, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8691601753234863, "num_tokens": 451654453.0, "step": 11837 }, { "epoch": 1.5059152779544587, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.506258964538574, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.867348313331604, "num_tokens": 451688302.0, "step": 11838 }, { "epoch": 1.5060424882330492, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.424019813537598, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8745777606964111, "num_tokens": 451727301.0, "step": 11839 }, { "epoch": 1.5061696985116397, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.614738464355469, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8546372652053833, "num_tokens": 451770471.0, "step": 11840 }, { "epoch": 1.5062969087902303, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.418803215026855, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8567472696304321, "num_tokens": 451809413.0, "step": 11841 }, { "epoch": 1.5064241190688208, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.500814437866211, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8658100366592407, "num_tokens": 451850177.0, "step": 11842 }, { "epoch": 1.5065513293474113, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.48433780670166, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.856523334980011, "num_tokens": 451887977.0, "step": 11843 }, { "epoch": 1.5066785396260016, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.56131649017334, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8790856003761292, "num_tokens": 451927403.0, "step": 11844 }, { "epoch": 1.5068057499045922, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.402599334716797, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8680613040924072, "num_tokens": 451967350.0, "step": 11845 }, { "epoch": 1.5069329601831827, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.514039039611816, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8683478236198425, "num_tokens": 452008664.0, "step": 11846 }, { "epoch": 1.5070601704617732, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.43167781829834, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8747944831848145, "num_tokens": 452047303.0, "step": 11847 }, { "epoch": 1.5071873807403637, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.4998140335083, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8706792593002319, "num_tokens": 452087359.0, "step": 11848 }, { "epoch": 1.5073145910189543, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512269973754883, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.854698657989502, "num_tokens": 452129426.0, "step": 11849 }, { "epoch": 1.5074418012975448, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.519262313842773, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8728796243667603, "num_tokens": 452169134.0, "step": 11850 }, { "epoch": 1.5075690115761353, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.542296409606934, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8586359024047852, "num_tokens": 452209962.0, "step": 11851 }, { "epoch": 1.5076962218547258, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.543224334716797, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8712932467460632, "num_tokens": 452245978.0, "step": 11852 }, { "epoch": 1.5078234321333164, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.599085807800293, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8778992295265198, "num_tokens": 452281013.0, "step": 11853 }, { "epoch": 1.507950642411907, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.54267692565918, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.882077157497406, "num_tokens": 452317174.0, "step": 11854 }, { "epoch": 1.5080778526904974, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.558095932006836, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8651354908943176, "num_tokens": 452362974.0, "step": 11855 }, { "epoch": 1.508205062969088, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.529460906982422, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8795229196548462, "num_tokens": 452400701.0, "step": 11856 }, { "epoch": 1.5083322732476785, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.53659725189209, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8780487179756165, "num_tokens": 452438541.0, "step": 11857 }, { "epoch": 1.508459483526269, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.535855293273926, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8632464408874512, "num_tokens": 452476918.0, "step": 11858 }, { "epoch": 1.5085866938048595, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.524757385253906, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8732562065124512, "num_tokens": 452509231.0, "step": 11859 }, { "epoch": 1.50871390408345, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.480568885803223, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8632690906524658, "num_tokens": 452547079.0, "step": 11860 }, { "epoch": 1.5088411143620406, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.488581657409668, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8786775469779968, "num_tokens": 452580215.0, "step": 11861 }, { "epoch": 1.508968324640631, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.545455932617188, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.874821662902832, "num_tokens": 452615778.0, "step": 11862 }, { "epoch": 1.5090955349192214, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.472171783447266, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8708595633506775, "num_tokens": 452652184.0, "step": 11863 }, { "epoch": 1.509222745197812, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512187004089355, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8546134233474731, "num_tokens": 452691345.0, "step": 11864 }, { "epoch": 1.5093499554764025, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.587438583374023, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8695147037506104, "num_tokens": 452729982.0, "step": 11865 }, { "epoch": 1.509477165754993, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.500255584716797, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8678485751152039, "num_tokens": 452771272.0, "step": 11866 }, { "epoch": 1.5096043760335836, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.500598907470703, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8645485639572144, "num_tokens": 452812231.0, "step": 11867 }, { "epoch": 1.509731586312174, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.58373737335205, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8576501607894897, "num_tokens": 452858406.0, "step": 11868 }, { "epoch": 1.5098587965907644, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.486977577209473, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8600029945373535, "num_tokens": 452898793.0, "step": 11869 }, { "epoch": 1.509986006869355, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.54006290435791, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8735297918319702, "num_tokens": 452937929.0, "step": 11870 }, { "epoch": 1.5101132171479454, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.486431121826172, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8712289929389954, "num_tokens": 452975436.0, "step": 11871 }, { "epoch": 1.510240427426536, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.51722526550293, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8693403601646423, "num_tokens": 453016334.0, "step": 11872 }, { "epoch": 1.5103676377051265, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.515547752380371, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8724046945571899, "num_tokens": 453053630.0, "step": 11873 }, { "epoch": 1.510494847983717, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.520341873168945, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.8507122993469238, "num_tokens": 453092341.0, "step": 11874 }, { "epoch": 1.5106220582623076, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.548434257507324, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8767744302749634, "num_tokens": 453126859.0, "step": 11875 }, { "epoch": 1.510749268540898, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.549074172973633, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.86539626121521, "num_tokens": 453167327.0, "step": 11876 }, { "epoch": 1.5108764788194886, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.456266403198242, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8746090531349182, "num_tokens": 453205901.0, "step": 11877 }, { "epoch": 1.5110036890980791, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.527843475341797, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8669837117195129, "num_tokens": 453241856.0, "step": 11878 }, { "epoch": 1.5111308993766697, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.48246955871582, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8769705295562744, "num_tokens": 453282528.0, "step": 11879 }, { "epoch": 1.5112581096552602, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.589071273803711, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8588470816612244, "num_tokens": 453318762.0, "step": 11880 }, { "epoch": 1.5113853199338507, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.520485877990723, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8837437033653259, "num_tokens": 453352482.0, "step": 11881 }, { "epoch": 1.5115125302124413, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.510821342468262, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8675193786621094, "num_tokens": 453393692.0, "step": 11882 }, { "epoch": 1.5116397404910318, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.541979789733887, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8727440237998962, "num_tokens": 453429217.0, "step": 11883 }, { "epoch": 1.5117669507696223, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.494696617126465, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8543092012405396, "num_tokens": 453468190.0, "step": 11884 }, { "epoch": 1.5118941610482128, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.550923347473145, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8620134592056274, "num_tokens": 453509893.0, "step": 11885 }, { "epoch": 1.5120213713268034, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.474324226379395, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8725666999816895, "num_tokens": 453546972.0, "step": 11886 }, { "epoch": 1.5121485816053937, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.540599822998047, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8802931308746338, "num_tokens": 453578480.0, "step": 11887 }, { "epoch": 1.5122757918839842, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.462732315063477, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8685020208358765, "num_tokens": 453623248.0, "step": 11888 }, { "epoch": 1.5124030021625747, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.457098007202148, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.888283371925354, "num_tokens": 453660088.0, "step": 11889 }, { "epoch": 1.5125302124411653, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.5386381149292, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8702813982963562, "num_tokens": 453698931.0, "step": 11890 }, { "epoch": 1.5126574227197558, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.518240928649902, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8708010911941528, "num_tokens": 453734220.0, "step": 11891 }, { "epoch": 1.5127846329983463, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.513087272644043, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8674600124359131, "num_tokens": 453774056.0, "step": 11892 }, { "epoch": 1.5129118432769366, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.463461875915527, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.875646710395813, "num_tokens": 453806616.0, "step": 11893 }, { "epoch": 1.5130390535555271, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.522382736206055, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8771370649337769, "num_tokens": 453843838.0, "step": 11894 }, { "epoch": 1.5131662638341177, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.495049476623535, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8761754035949707, "num_tokens": 453876888.0, "step": 11895 }, { "epoch": 1.5132934741127082, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.52490520477295, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8840569257736206, "num_tokens": 453915334.0, "step": 11896 }, { "epoch": 1.5134206843912987, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.52174186706543, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8788504600524902, "num_tokens": 453950590.0, "step": 11897 }, { "epoch": 1.5135478946698893, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.486225128173828, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8729460835456848, "num_tokens": 453992590.0, "step": 11898 }, { "epoch": 1.5136751049484798, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.530670166015625, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8734843134880066, "num_tokens": 454038619.0, "step": 11899 }, { "epoch": 1.5138023152270703, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.553393363952637, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8602652549743652, "num_tokens": 454081564.0, "step": 11900 }, { "epoch": 1.5139295255056608, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.413229942321777, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8564913272857666, "num_tokens": 454125581.0, "step": 11901 }, { "epoch": 1.5140567357842514, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.531611442565918, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8780398964881897, "num_tokens": 454163961.0, "step": 11902 }, { "epoch": 1.514183946062842, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.508903503417969, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8621361255645752, "num_tokens": 454200410.0, "step": 11903 }, { "epoch": 1.5143111563414324, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.511588096618652, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8549231886863708, "num_tokens": 454237143.0, "step": 11904 }, { "epoch": 1.514438366620023, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.533140182495117, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8759934306144714, "num_tokens": 454274022.0, "step": 11905 }, { "epoch": 1.5145655768986135, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.483698844909668, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8636065721511841, "num_tokens": 454307955.0, "step": 11906 }, { "epoch": 1.514692787177204, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.550338745117188, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8761889934539795, "num_tokens": 454345449.0, "step": 11907 }, { "epoch": 1.5148199974557945, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.468098640441895, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8577171564102173, "num_tokens": 454383096.0, "step": 11908 }, { "epoch": 1.514947207734385, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.553309440612793, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.857994019985199, "num_tokens": 454420142.0, "step": 11909 }, { "epoch": 1.5150744180129756, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.48848819732666, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8677534461021423, "num_tokens": 454454991.0, "step": 11910 }, { "epoch": 1.515201628291566, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.509026527404785, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.865045964717865, "num_tokens": 454490161.0, "step": 11911 }, { "epoch": 1.5153288385701564, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.498772621154785, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.860450267791748, "num_tokens": 454534767.0, "step": 11912 }, { "epoch": 1.515456048848747, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.516758918762207, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8778005838394165, "num_tokens": 454573246.0, "step": 11913 }, { "epoch": 1.5155832591273375, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.511494636535645, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.881608247756958, "num_tokens": 454607626.0, "step": 11914 }, { "epoch": 1.515710469405928, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.52548599243164, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8621663451194763, "num_tokens": 454639840.0, "step": 11915 }, { "epoch": 1.5158376796845185, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.49471378326416, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8602689504623413, "num_tokens": 454681445.0, "step": 11916 }, { "epoch": 1.515964889963109, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.463850975036621, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8617982864379883, "num_tokens": 454726191.0, "step": 11917 }, { "epoch": 1.5160921002416994, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.524507522583008, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8754629492759705, "num_tokens": 454765112.0, "step": 11918 }, { "epoch": 1.51621931052029, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.507107734680176, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8668445348739624, "num_tokens": 454805254.0, "step": 11919 }, { "epoch": 1.5163465207988804, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.599794387817383, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8613587617874146, "num_tokens": 454841344.0, "step": 11920 }, { "epoch": 1.516473731077471, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.492259979248047, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8821507692337036, "num_tokens": 454879986.0, "step": 11921 }, { "epoch": 1.5166009413560615, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.574175834655762, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8603709936141968, "num_tokens": 454917731.0, "step": 11922 }, { "epoch": 1.516728151634652, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.465790748596191, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8741279244422913, "num_tokens": 454952837.0, "step": 11923 }, { "epoch": 1.5168553619132426, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.583033561706543, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8719769716262817, "num_tokens": 454992751.0, "step": 11924 }, { "epoch": 1.516982572191833, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.538987159729004, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8672648072242737, "num_tokens": 455029829.0, "step": 11925 }, { "epoch": 1.5171097824704236, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.514309883117676, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8695228695869446, "num_tokens": 455072167.0, "step": 11926 }, { "epoch": 1.5172369927490141, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.53276252746582, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.876264214515686, "num_tokens": 455108433.0, "step": 11927 }, { "epoch": 1.5173642030276047, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.516786575317383, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8720831274986267, "num_tokens": 455148943.0, "step": 11928 }, { "epoch": 1.5174914133061952, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.553296089172363, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8646200895309448, "num_tokens": 455185099.0, "step": 11929 }, { "epoch": 1.5176186235847857, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.511231422424316, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8829944729804993, "num_tokens": 455225322.0, "step": 11930 }, { "epoch": 1.5177458338633762, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.550445556640625, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8656471967697144, "num_tokens": 455262830.0, "step": 11931 }, { "epoch": 1.5178730441419668, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.521280288696289, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8754538297653198, "num_tokens": 455300825.0, "step": 11932 }, { "epoch": 1.5180002544205573, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.583300590515137, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8657290935516357, "num_tokens": 455336209.0, "step": 11933 }, { "epoch": 1.5181274646991478, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.464223861694336, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8682229518890381, "num_tokens": 455374909.0, "step": 11934 }, { "epoch": 1.5182546749777384, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.54434871673584, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8814102411270142, "num_tokens": 455417179.0, "step": 11935 }, { "epoch": 1.5183818852563287, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.490798950195312, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8557102680206299, "num_tokens": 455457923.0, "step": 11936 }, { "epoch": 1.5185090955349192, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.512550354003906, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8600785732269287, "num_tokens": 455492932.0, "step": 11937 }, { "epoch": 1.5186363058135097, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.501121520996094, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8598968386650085, "num_tokens": 455540392.0, "step": 11938 }, { "epoch": 1.5187635160921003, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.559333801269531, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8642064332962036, "num_tokens": 455574461.0, "step": 11939 }, { "epoch": 1.5188907263706908, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.541275978088379, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8656432628631592, "num_tokens": 455608315.0, "step": 11940 }, { "epoch": 1.5190179366492813, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.51648235321045, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8727885484695435, "num_tokens": 455654927.0, "step": 11941 }, { "epoch": 1.5191451469278716, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.534891128540039, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8563677668571472, "num_tokens": 455695745.0, "step": 11942 }, { "epoch": 1.5192723572064621, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.491242408752441, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8695290684700012, "num_tokens": 455734652.0, "step": 11943 }, { "epoch": 1.5193995674850527, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.520845413208008, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8466518521308899, "num_tokens": 455776063.0, "step": 11944 }, { "epoch": 1.5195267777636432, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.55705451965332, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.863503098487854, "num_tokens": 455820595.0, "step": 11945 }, { "epoch": 1.5196539880422337, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.541142463684082, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8744301199913025, "num_tokens": 455859553.0, "step": 11946 }, { "epoch": 1.5197811983208243, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.431817054748535, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8689954876899719, "num_tokens": 455892676.0, "step": 11947 }, { "epoch": 1.5199084085994148, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.481612205505371, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8607984781265259, "num_tokens": 455932106.0, "step": 11948 }, { "epoch": 1.5200356188780053, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.513172149658203, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.873675525188446, "num_tokens": 455971555.0, "step": 11949 }, { "epoch": 1.5201628291565958, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.462325096130371, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8635708093643188, "num_tokens": 456014177.0, "step": 11950 }, { "epoch": 1.5202900394351864, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.571793556213379, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8739041090011597, "num_tokens": 456048229.0, "step": 11951 }, { "epoch": 1.520417249713777, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.435819625854492, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.887088418006897, "num_tokens": 456085815.0, "step": 11952 }, { "epoch": 1.5205444599923674, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.508803367614746, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8708500266075134, "num_tokens": 456128971.0, "step": 11953 }, { "epoch": 1.520671670270958, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.481757164001465, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8519574403762817, "num_tokens": 456159705.0, "step": 11954 }, { "epoch": 1.5207988805495485, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.559844017028809, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8597241044044495, "num_tokens": 456198251.0, "step": 11955 }, { "epoch": 1.520926090828139, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.517949104309082, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8612638711929321, "num_tokens": 456237612.0, "step": 11956 }, { "epoch": 1.5210533011067295, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.543639183044434, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8679773807525635, "num_tokens": 456273665.0, "step": 11957 }, { "epoch": 1.52118051138532, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.519957542419434, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8697463274002075, "num_tokens": 456305467.0, "step": 11958 }, { "epoch": 1.5213077216639106, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.472250938415527, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8752688765525818, "num_tokens": 456342951.0, "step": 11959 }, { "epoch": 1.521434931942501, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.477845191955566, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8682562112808228, "num_tokens": 456375846.0, "step": 11960 }, { "epoch": 1.5215621422210914, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.512353897094727, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8777374029159546, "num_tokens": 456407993.0, "step": 11961 }, { "epoch": 1.521689352499682, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.515217781066895, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8497785329818726, "num_tokens": 456444278.0, "step": 11962 }, { "epoch": 1.5218165627782725, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.504382133483887, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.877769947052002, "num_tokens": 456483950.0, "step": 11963 }, { "epoch": 1.521943773056863, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.505252838134766, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8693726062774658, "num_tokens": 456527438.0, "step": 11964 }, { "epoch": 1.5220709833354535, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.504965782165527, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8785557746887207, "num_tokens": 456560572.0, "step": 11965 }, { "epoch": 1.5221981936140438, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.527804374694824, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8598862886428833, "num_tokens": 456604781.0, "step": 11966 }, { "epoch": 1.5223254038926344, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.503424644470215, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8824887871742249, "num_tokens": 456639415.0, "step": 11967 }, { "epoch": 1.522452614171225, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.522826194763184, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8810008764266968, "num_tokens": 456671183.0, "step": 11968 }, { "epoch": 1.5225798244498154, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.504767417907715, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8570390939712524, "num_tokens": 456712995.0, "step": 11969 }, { "epoch": 1.522707034728406, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.489004135131836, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8642966747283936, "num_tokens": 456750587.0, "step": 11970 }, { "epoch": 1.5228342450069965, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.535428047180176, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8762210607528687, "num_tokens": 456787503.0, "step": 11971 }, { "epoch": 1.522961455285587, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.493252754211426, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8470775485038757, "num_tokens": 456831645.0, "step": 11972 }, { "epoch": 1.5230886655641775, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.502242088317871, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8755698204040527, "num_tokens": 456869143.0, "step": 11973 }, { "epoch": 1.523215875842768, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.60680866241455, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.866862952709198, "num_tokens": 456906465.0, "step": 11974 }, { "epoch": 1.5233430861213586, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.46491527557373, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8678272366523743, "num_tokens": 456942048.0, "step": 11975 }, { "epoch": 1.5234702963999491, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.606574058532715, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8865118622779846, "num_tokens": 456975710.0, "step": 11976 }, { "epoch": 1.5235975066785397, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.456660270690918, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8579674959182739, "num_tokens": 457022295.0, "step": 11977 }, { "epoch": 1.5237247169571302, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.581955909729004, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.861291766166687, "num_tokens": 457060433.0, "step": 11978 }, { "epoch": 1.5238519272357207, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.545095443725586, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8747882843017578, "num_tokens": 457101760.0, "step": 11979 }, { "epoch": 1.5239791375143112, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.526507377624512, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8848080635070801, "num_tokens": 457142676.0, "step": 11980 }, { "epoch": 1.5241063477929018, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.508872032165527, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8757251501083374, "num_tokens": 457179507.0, "step": 11981 }, { "epoch": 1.5242335580714923, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.455031394958496, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8699744939804077, "num_tokens": 457221563.0, "step": 11982 }, { "epoch": 1.5243607683500828, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.505382537841797, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8665094375610352, "num_tokens": 457258634.0, "step": 11983 }, { "epoch": 1.5244879786286734, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.471341133117676, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8828467130661011, "num_tokens": 457293097.0, "step": 11984 }, { "epoch": 1.5246151889072637, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.478991508483887, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8834857940673828, "num_tokens": 457333525.0, "step": 11985 }, { "epoch": 1.5247423991858542, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512118339538574, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.854529082775116, "num_tokens": 457379772.0, "step": 11986 }, { "epoch": 1.5248696094644447, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.477865219116211, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8749259114265442, "num_tokens": 457415852.0, "step": 11987 }, { "epoch": 1.5249968197430352, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.545649528503418, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8735359907150269, "num_tokens": 457449771.0, "step": 11988 }, { "epoch": 1.5251240300216258, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.508861541748047, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8648436069488525, "num_tokens": 457489601.0, "step": 11989 }, { "epoch": 1.5252512403002163, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.540343284606934, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8833679556846619, "num_tokens": 457530516.0, "step": 11990 }, { "epoch": 1.5253784505788066, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.506025314331055, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8655738830566406, "num_tokens": 457568299.0, "step": 11991 }, { "epoch": 1.5255056608573971, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.488258361816406, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8682499527931213, "num_tokens": 457600476.0, "step": 11992 }, { "epoch": 1.5256328711359877, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.540410995483398, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.88933265209198, "num_tokens": 457632024.0, "step": 11993 }, { "epoch": 1.5257600814145782, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.454222679138184, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8742274045944214, "num_tokens": 457670805.0, "step": 11994 }, { "epoch": 1.5258872916931687, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.52172565460205, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8703822493553162, "num_tokens": 457702468.0, "step": 11995 }, { "epoch": 1.5260145019717593, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.480936050415039, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8782299757003784, "num_tokens": 457740443.0, "step": 11996 }, { "epoch": 1.5261417122503498, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.46730899810791, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.870734453201294, "num_tokens": 457775833.0, "step": 11997 }, { "epoch": 1.5262689225289403, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.534420013427734, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8730552196502686, "num_tokens": 457815850.0, "step": 11998 }, { "epoch": 1.5263961328075308, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.508357048034668, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8766512870788574, "num_tokens": 457854946.0, "step": 11999 }, { "epoch": 1.5265233430861214, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.527138710021973, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8741458058357239, "num_tokens": 457890775.0, "step": 12000 }, { "epoch": 1.526650553364712, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.491405487060547, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8749189972877502, "num_tokens": 457928143.0, "step": 12001 }, { "epoch": 1.5267777636433024, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.496814727783203, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8758929371833801, "num_tokens": 457965999.0, "step": 12002 }, { "epoch": 1.526904973921893, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.525294303894043, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.86965012550354, "num_tokens": 458006638.0, "step": 12003 }, { "epoch": 1.5270321842004835, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.471603393554688, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8703579902648926, "num_tokens": 458047522.0, "step": 12004 }, { "epoch": 1.527159394479074, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.480595588684082, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.869433581829071, "num_tokens": 458091219.0, "step": 12005 }, { "epoch": 1.5272866047576645, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.461702346801758, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8586198091506958, "num_tokens": 458129208.0, "step": 12006 }, { "epoch": 1.527413815036255, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.509793281555176, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8644728660583496, "num_tokens": 458163632.0, "step": 12007 }, { "epoch": 1.5275410253148456, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.551875114440918, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8632698059082031, "num_tokens": 458202377.0, "step": 12008 }, { "epoch": 1.527668235593436, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.469802856445312, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8704747557640076, "num_tokens": 458239901.0, "step": 12009 }, { "epoch": 1.5277954458720264, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.606167793273926, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8644062280654907, "num_tokens": 458280291.0, "step": 12010 }, { "epoch": 1.527922656150617, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.531193733215332, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8751118779182434, "num_tokens": 458316931.0, "step": 12011 }, { "epoch": 1.5280498664292075, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.585577964782715, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8629080057144165, "num_tokens": 458356477.0, "step": 12012 }, { "epoch": 1.528177076707798, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.507453918457031, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8557112216949463, "num_tokens": 458401340.0, "step": 12013 }, { "epoch": 1.5283042869863885, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.585161209106445, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8763208389282227, "num_tokens": 458436190.0, "step": 12014 }, { "epoch": 1.5284314972649788, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.515043258666992, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8694281578063965, "num_tokens": 458465513.0, "step": 12015 }, { "epoch": 1.5285587075435694, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.563228607177734, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8657633066177368, "num_tokens": 458505465.0, "step": 12016 }, { "epoch": 1.52868591782216, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.527531623840332, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8673976063728333, "num_tokens": 458545087.0, "step": 12017 }, { "epoch": 1.5288131281007504, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.505452156066895, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8773183822631836, "num_tokens": 458583274.0, "step": 12018 }, { "epoch": 1.528940338379341, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.496671676635742, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8627015948295593, "num_tokens": 458626026.0, "step": 12019 }, { "epoch": 1.5290675486579315, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.550993919372559, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8806561231613159, "num_tokens": 458668774.0, "step": 12020 }, { "epoch": 1.529194758936522, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.528077125549316, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8729153871536255, "num_tokens": 458702256.0, "step": 12021 }, { "epoch": 1.5293219692151125, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.52807903289795, "learning_rate": 1e-06, "loss": 0.4932, "mean_token_accuracy": 0.8424925804138184, "num_tokens": 458741429.0, "step": 12022 }, { "epoch": 1.529449179493703, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.536698341369629, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8743835687637329, "num_tokens": 458775608.0, "step": 12023 }, { "epoch": 1.5295763897722936, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.534318923950195, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8730915188789368, "num_tokens": 458811451.0, "step": 12024 }, { "epoch": 1.5297036000508841, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.501842498779297, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8736134171485901, "num_tokens": 458851761.0, "step": 12025 }, { "epoch": 1.5298308103294747, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.522342681884766, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.863017201423645, "num_tokens": 458888948.0, "step": 12026 }, { "epoch": 1.5299580206080652, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.582331657409668, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8669817447662354, "num_tokens": 458922466.0, "step": 12027 }, { "epoch": 1.5300852308866557, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.549654960632324, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8651334643363953, "num_tokens": 458956947.0, "step": 12028 }, { "epoch": 1.5302124411652462, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53611946105957, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.887223482131958, "num_tokens": 458993634.0, "step": 12029 }, { "epoch": 1.5303396514438368, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.55045223236084, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8780953884124756, "num_tokens": 459034314.0, "step": 12030 }, { "epoch": 1.5304668617224273, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.58430004119873, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8747788667678833, "num_tokens": 459072117.0, "step": 12031 }, { "epoch": 1.5305940720010178, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.632707595825195, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8645966649055481, "num_tokens": 459112382.0, "step": 12032 }, { "epoch": 1.5307212822796084, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.503656387329102, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8822751045227051, "num_tokens": 459148973.0, "step": 12033 }, { "epoch": 1.5308484925581987, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.606461524963379, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8672395944595337, "num_tokens": 459183242.0, "step": 12034 }, { "epoch": 1.5309757028367892, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.52450942993164, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.8862110376358032, "num_tokens": 459219606.0, "step": 12035 }, { "epoch": 1.5311029131153797, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.556907653808594, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8881859183311462, "num_tokens": 459253763.0, "step": 12036 }, { "epoch": 1.5312301233939702, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.542250633239746, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8734356164932251, "num_tokens": 459291187.0, "step": 12037 }, { "epoch": 1.5313573336725608, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.533447265625, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8669002652168274, "num_tokens": 459327423.0, "step": 12038 }, { "epoch": 1.5314845439511513, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.546607971191406, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8664619326591492, "num_tokens": 459366207.0, "step": 12039 }, { "epoch": 1.5316117542297416, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.525629997253418, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8614394664764404, "num_tokens": 459408338.0, "step": 12040 }, { "epoch": 1.5317389645083321, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.55334186553955, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8801544904708862, "num_tokens": 459443420.0, "step": 12041 }, { "epoch": 1.5318661747869227, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.576178550720215, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8542835712432861, "num_tokens": 459475835.0, "step": 12042 }, { "epoch": 1.5319933850655132, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.56391429901123, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8682321310043335, "num_tokens": 459511993.0, "step": 12043 }, { "epoch": 1.5321205953441037, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.585766792297363, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8771939873695374, "num_tokens": 459546221.0, "step": 12044 }, { "epoch": 1.5322478056226942, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.529107093811035, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8742465972900391, "num_tokens": 459587271.0, "step": 12045 }, { "epoch": 1.5323750159012848, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.5496826171875, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8593990206718445, "num_tokens": 459623388.0, "step": 12046 }, { "epoch": 1.5325022261798753, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.606078147888184, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8684796094894409, "num_tokens": 459666440.0, "step": 12047 }, { "epoch": 1.5326294364584658, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.519379615783691, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8579154014587402, "num_tokens": 459708373.0, "step": 12048 }, { "epoch": 1.5327566467370564, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.570862770080566, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8799881935119629, "num_tokens": 459746707.0, "step": 12049 }, { "epoch": 1.532883857015647, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.470739364624023, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8757848739624023, "num_tokens": 459792493.0, "step": 12050 }, { "epoch": 1.5330110672942374, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.514689445495605, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8787292242050171, "num_tokens": 459828897.0, "step": 12051 }, { "epoch": 1.533138277572828, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.558096885681152, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8618623614311218, "num_tokens": 459866274.0, "step": 12052 }, { "epoch": 1.5332654878514185, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.568127632141113, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8854711055755615, "num_tokens": 459902442.0, "step": 12053 }, { "epoch": 1.533392698130009, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.4913911819458, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8778182864189148, "num_tokens": 459943331.0, "step": 12054 }, { "epoch": 1.5335199084085995, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.594837188720703, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.875514566898346, "num_tokens": 459975349.0, "step": 12055 }, { "epoch": 1.53364711868719, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.588580131530762, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8834272623062134, "num_tokens": 460008295.0, "step": 12056 }, { "epoch": 1.5337743289657806, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.603352546691895, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8749931454658508, "num_tokens": 460045892.0, "step": 12057 }, { "epoch": 1.533901539244371, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512804985046387, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8704047203063965, "num_tokens": 460084255.0, "step": 12058 }, { "epoch": 1.5340287495229614, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.579133033752441, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8566805124282837, "num_tokens": 460123706.0, "step": 12059 }, { "epoch": 1.534155959801552, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.539076805114746, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8562780022621155, "num_tokens": 460162351.0, "step": 12060 }, { "epoch": 1.5342831700801425, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.54623794555664, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8605508804321289, "num_tokens": 460200135.0, "step": 12061 }, { "epoch": 1.534410380358733, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.562665939331055, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8637410998344421, "num_tokens": 460243585.0, "step": 12062 }, { "epoch": 1.5345375906373235, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.514738082885742, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8640869855880737, "num_tokens": 460283695.0, "step": 12063 }, { "epoch": 1.5346648009159138, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.572794914245605, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8773294687271118, "num_tokens": 460317241.0, "step": 12064 }, { "epoch": 1.5347920111945044, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.58579158782959, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8721718192100525, "num_tokens": 460351900.0, "step": 12065 }, { "epoch": 1.534919221473095, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.479792594909668, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8724145889282227, "num_tokens": 460391672.0, "step": 12066 }, { "epoch": 1.5350464317516854, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.606562614440918, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8623290061950684, "num_tokens": 460434120.0, "step": 12067 }, { "epoch": 1.535173642030276, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.59789752960205, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8715043067932129, "num_tokens": 460474214.0, "step": 12068 }, { "epoch": 1.5353008523088665, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.560242652893066, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8760569095611572, "num_tokens": 460507549.0, "step": 12069 }, { "epoch": 1.535428062587457, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.611202239990234, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8631553649902344, "num_tokens": 460546391.0, "step": 12070 }, { "epoch": 1.5355552728660475, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.556000709533691, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8590482473373413, "num_tokens": 460585232.0, "step": 12071 }, { "epoch": 1.535682483144638, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.588571548461914, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8782631158828735, "num_tokens": 460624398.0, "step": 12072 }, { "epoch": 1.5358096934232286, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.538435935974121, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8632686138153076, "num_tokens": 460667193.0, "step": 12073 }, { "epoch": 1.5359369037018191, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.574557304382324, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8467177152633667, "num_tokens": 460710614.0, "step": 12074 }, { "epoch": 1.5360641139804097, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.563651084899902, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8660525679588318, "num_tokens": 460751079.0, "step": 12075 }, { "epoch": 1.5361913242590002, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.59329891204834, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8696110248565674, "num_tokens": 460786768.0, "step": 12076 }, { "epoch": 1.5363185345375907, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.609633445739746, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8604822158813477, "num_tokens": 460824757.0, "step": 12077 }, { "epoch": 1.5364457448161812, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.604960441589355, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8631073832511902, "num_tokens": 460865024.0, "step": 12078 }, { "epoch": 1.5365729550947718, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.601076126098633, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8606545925140381, "num_tokens": 460900895.0, "step": 12079 }, { "epoch": 1.5367001653733623, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.546467781066895, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8754584789276123, "num_tokens": 460936876.0, "step": 12080 }, { "epoch": 1.5368273756519528, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.549599647521973, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8546028137207031, "num_tokens": 460975297.0, "step": 12081 }, { "epoch": 1.5369545859305433, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.576805114746094, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8583531379699707, "num_tokens": 461008603.0, "step": 12082 }, { "epoch": 1.5370817962091337, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.538981437683105, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8608372211456299, "num_tokens": 461047770.0, "step": 12083 }, { "epoch": 1.5372090064877242, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.552292823791504, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8571798801422119, "num_tokens": 461088554.0, "step": 12084 }, { "epoch": 1.5373362167663147, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.51774787902832, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8482600450515747, "num_tokens": 461126070.0, "step": 12085 }, { "epoch": 1.5374634270449052, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.5927095413208, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8668254017829895, "num_tokens": 461166394.0, "step": 12086 }, { "epoch": 1.5375906373234958, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.509879112243652, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8623496890068054, "num_tokens": 461208263.0, "step": 12087 }, { "epoch": 1.5377178476020863, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.583821296691895, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8792601227760315, "num_tokens": 461245054.0, "step": 12088 }, { "epoch": 1.5378450578806766, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.532711029052734, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8583972454071045, "num_tokens": 461281584.0, "step": 12089 }, { "epoch": 1.5379722681592671, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.529454231262207, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8779471516609192, "num_tokens": 461322247.0, "step": 12090 }, { "epoch": 1.5380994784378577, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.587403297424316, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.857061505317688, "num_tokens": 461358371.0, "step": 12091 }, { "epoch": 1.5382266887164482, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.492823600769043, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8751159906387329, "num_tokens": 461394666.0, "step": 12092 }, { "epoch": 1.5383538989950387, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.584336280822754, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8807330131530762, "num_tokens": 461430875.0, "step": 12093 }, { "epoch": 1.5384811092736292, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.526204109191895, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8659922480583191, "num_tokens": 461467412.0, "step": 12094 }, { "epoch": 1.5386083195522198, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.583292007446289, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8483975529670715, "num_tokens": 461501970.0, "step": 12095 }, { "epoch": 1.5387355298308103, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.55998706817627, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8569196462631226, "num_tokens": 461542735.0, "step": 12096 }, { "epoch": 1.5388627401094008, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.598811149597168, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8710336089134216, "num_tokens": 461583883.0, "step": 12097 }, { "epoch": 1.5389899503879914, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.550999641418457, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8615754842758179, "num_tokens": 461625124.0, "step": 12098 }, { "epoch": 1.5391171606665819, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.611313819885254, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.859445333480835, "num_tokens": 461660273.0, "step": 12099 }, { "epoch": 1.5392443709451724, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.60055160522461, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.864006519317627, "num_tokens": 461698890.0, "step": 12100 }, { "epoch": 1.539371581223763, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.648419380187988, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8684159517288208, "num_tokens": 461736927.0, "step": 12101 }, { "epoch": 1.5394987915023535, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.474676132202148, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8665539026260376, "num_tokens": 461777796.0, "step": 12102 }, { "epoch": 1.539626001780944, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.687499046325684, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8786306381225586, "num_tokens": 461814656.0, "step": 12103 }, { "epoch": 1.5397532120595345, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.517168045043945, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8513385057449341, "num_tokens": 461852186.0, "step": 12104 }, { "epoch": 1.539880422338125, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.556821823120117, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8816545009613037, "num_tokens": 461893724.0, "step": 12105 }, { "epoch": 1.5400076326167156, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.569894790649414, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8559111952781677, "num_tokens": 461935380.0, "step": 12106 }, { "epoch": 1.5401348428953059, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.533243179321289, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.874196469783783, "num_tokens": 461971463.0, "step": 12107 }, { "epoch": 1.5402620531738964, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.584101676940918, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8753743171691895, "num_tokens": 462007514.0, "step": 12108 }, { "epoch": 1.540389263452487, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.525360107421875, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8820421695709229, "num_tokens": 462046022.0, "step": 12109 }, { "epoch": 1.5405164737310775, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.576135635375977, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.873257577419281, "num_tokens": 462089367.0, "step": 12110 }, { "epoch": 1.540643684009668, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.562219619750977, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8743224143981934, "num_tokens": 462128558.0, "step": 12111 }, { "epoch": 1.5407708942882585, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.500726699829102, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8758821487426758, "num_tokens": 462162639.0, "step": 12112 }, { "epoch": 1.5408981045668488, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.582369804382324, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8697107434272766, "num_tokens": 462199303.0, "step": 12113 }, { "epoch": 1.5410253148454394, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.536544799804688, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8673326969146729, "num_tokens": 462246095.0, "step": 12114 }, { "epoch": 1.54115252512403, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.621797561645508, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8658173084259033, "num_tokens": 462286653.0, "step": 12115 }, { "epoch": 1.5412797354026204, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.697022438049316, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8654148578643799, "num_tokens": 462328295.0, "step": 12116 }, { "epoch": 1.541406945681211, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.568976402282715, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8708802461624146, "num_tokens": 462365175.0, "step": 12117 }, { "epoch": 1.5415341559598015, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.542585372924805, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8616054654121399, "num_tokens": 462407044.0, "step": 12118 }, { "epoch": 1.541661366238392, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.59821891784668, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8626048564910889, "num_tokens": 462443819.0, "step": 12119 }, { "epoch": 1.5417885765169825, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.55231761932373, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8782129883766174, "num_tokens": 462487319.0, "step": 12120 }, { "epoch": 1.541915786795573, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.550079345703125, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8698742389678955, "num_tokens": 462527026.0, "step": 12121 }, { "epoch": 1.5420429970741636, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.634109497070312, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8598313927650452, "num_tokens": 462557552.0, "step": 12122 }, { "epoch": 1.5421702073527541, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.581165313720703, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8704519867897034, "num_tokens": 462595243.0, "step": 12123 }, { "epoch": 1.5422974176313446, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.561960220336914, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8712610006332397, "num_tokens": 462636077.0, "step": 12124 }, { "epoch": 1.5424246279099352, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.610305786132812, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8801808953285217, "num_tokens": 462668407.0, "step": 12125 }, { "epoch": 1.5425518381885257, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.552088737487793, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8690363168716431, "num_tokens": 462707678.0, "step": 12126 }, { "epoch": 1.5426790484671162, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.566254615783691, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8615977168083191, "num_tokens": 462739722.0, "step": 12127 }, { "epoch": 1.5428062587457068, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.641634941101074, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8764592409133911, "num_tokens": 462770369.0, "step": 12128 }, { "epoch": 1.5429334690242973, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.564482688903809, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8752733469009399, "num_tokens": 462806126.0, "step": 12129 }, { "epoch": 1.5430606793028878, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.533073425292969, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8705630302429199, "num_tokens": 462846585.0, "step": 12130 }, { "epoch": 1.5431878895814783, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.541860580444336, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8655799627304077, "num_tokens": 462889892.0, "step": 12131 }, { "epoch": 1.5433150998600687, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.596209526062012, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8687006235122681, "num_tokens": 462927612.0, "step": 12132 }, { "epoch": 1.5434423101386592, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.547114372253418, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.87115478515625, "num_tokens": 462971661.0, "step": 12133 }, { "epoch": 1.5435695204172497, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.584959983825684, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8739333152770996, "num_tokens": 463011260.0, "step": 12134 }, { "epoch": 1.5436967306958402, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.58174991607666, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8759211301803589, "num_tokens": 463045395.0, "step": 12135 }, { "epoch": 1.5438239409744308, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.553079605102539, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8589258790016174, "num_tokens": 463081015.0, "step": 12136 }, { "epoch": 1.5439511512530213, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.630284309387207, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8682308197021484, "num_tokens": 463115658.0, "step": 12137 }, { "epoch": 1.5440783615316116, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.531506538391113, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8817015886306763, "num_tokens": 463152084.0, "step": 12138 }, { "epoch": 1.5442055718102021, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.49779224395752, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8595174551010132, "num_tokens": 463190137.0, "step": 12139 }, { "epoch": 1.5443327820887927, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.568717002868652, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8688201308250427, "num_tokens": 463227626.0, "step": 12140 }, { "epoch": 1.5444599923673832, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.514968872070312, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8695178031921387, "num_tokens": 463266451.0, "step": 12141 }, { "epoch": 1.5445872026459737, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.699542045593262, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.857384979724884, "num_tokens": 463305327.0, "step": 12142 }, { "epoch": 1.5447144129245642, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.526670455932617, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8654187917709351, "num_tokens": 463345890.0, "step": 12143 }, { "epoch": 1.5448416232031548, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.542913436889648, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8735729455947876, "num_tokens": 463381583.0, "step": 12144 }, { "epoch": 1.5449688334817453, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.539568901062012, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8774398565292358, "num_tokens": 463414442.0, "step": 12145 }, { "epoch": 1.5450960437603358, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.639325141906738, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8753832578659058, "num_tokens": 463449517.0, "step": 12146 }, { "epoch": 1.5452232540389264, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.513937950134277, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8756941556930542, "num_tokens": 463487952.0, "step": 12147 }, { "epoch": 1.5453504643175169, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.631535530090332, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.869568943977356, "num_tokens": 463520724.0, "step": 12148 }, { "epoch": 1.5454776745961074, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.557193756103516, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8663815855979919, "num_tokens": 463564691.0, "step": 12149 }, { "epoch": 1.545604884874698, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.603577613830566, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8656628727912903, "num_tokens": 463599037.0, "step": 12150 }, { "epoch": 1.5457320951532885, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.697671890258789, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8566496968269348, "num_tokens": 463630744.0, "step": 12151 }, { "epoch": 1.545859305431879, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.588696479797363, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8652406930923462, "num_tokens": 463663551.0, "step": 12152 }, { "epoch": 1.5459865157104695, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.551265716552734, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8807076811790466, "num_tokens": 463702446.0, "step": 12153 }, { "epoch": 1.54611372598906, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.577457427978516, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8608794212341309, "num_tokens": 463739324.0, "step": 12154 }, { "epoch": 1.5462409362676506, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.58573055267334, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8497624397277832, "num_tokens": 463780765.0, "step": 12155 }, { "epoch": 1.5463681465462409, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.606557846069336, "learning_rate": 1e-06, "loss": 0.4926, "mean_token_accuracy": 0.8476351499557495, "num_tokens": 463817727.0, "step": 12156 }, { "epoch": 1.5464953568248314, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.551939964294434, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8833197355270386, "num_tokens": 463865793.0, "step": 12157 }, { "epoch": 1.546622567103422, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.580025672912598, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8699360489845276, "num_tokens": 463908299.0, "step": 12158 }, { "epoch": 1.5467497773820125, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.571573257446289, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8660668730735779, "num_tokens": 463944513.0, "step": 12159 }, { "epoch": 1.546876987660603, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.537535667419434, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8620356321334839, "num_tokens": 463982862.0, "step": 12160 }, { "epoch": 1.5470041979391935, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.572898864746094, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8734855055809021, "num_tokens": 464023357.0, "step": 12161 }, { "epoch": 1.5471314082177838, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.56109619140625, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8685606122016907, "num_tokens": 464071889.0, "step": 12162 }, { "epoch": 1.5472586184963744, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.564818382263184, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8560236692428589, "num_tokens": 464112334.0, "step": 12163 }, { "epoch": 1.5473858287749649, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.572858810424805, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8644506335258484, "num_tokens": 464156833.0, "step": 12164 }, { "epoch": 1.5475130390535554, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.528429985046387, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8721619248390198, "num_tokens": 464192989.0, "step": 12165 }, { "epoch": 1.547640249332146, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.564749717712402, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.875495195388794, "num_tokens": 464232855.0, "step": 12166 }, { "epoch": 1.5477674596107365, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.622345924377441, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8531258702278137, "num_tokens": 464271860.0, "step": 12167 }, { "epoch": 1.547894669889327, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512958526611328, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8816161155700684, "num_tokens": 464312925.0, "step": 12168 }, { "epoch": 1.5480218801679175, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.562793731689453, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8636929988861084, "num_tokens": 464349190.0, "step": 12169 }, { "epoch": 1.548149090446508, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.57142162322998, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8644857406616211, "num_tokens": 464392031.0, "step": 12170 }, { "epoch": 1.5482763007250986, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.524060249328613, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8652209043502808, "num_tokens": 464424884.0, "step": 12171 }, { "epoch": 1.5484035110036891, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.620575904846191, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8633622527122498, "num_tokens": 464466047.0, "step": 12172 }, { "epoch": 1.5485307212822796, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.572766304016113, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8668244481086731, "num_tokens": 464497293.0, "step": 12173 }, { "epoch": 1.5486579315608702, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.557695388793945, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8550046682357788, "num_tokens": 464541873.0, "step": 12174 }, { "epoch": 1.5487851418394607, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.629424095153809, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8696177005767822, "num_tokens": 464579712.0, "step": 12175 }, { "epoch": 1.5489123521180512, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.597529411315918, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.867039680480957, "num_tokens": 464612277.0, "step": 12176 }, { "epoch": 1.5490395623966418, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.620234489440918, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8655377626419067, "num_tokens": 464647062.0, "step": 12177 }, { "epoch": 1.5491667726752323, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.594882011413574, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8707619905471802, "num_tokens": 464677998.0, "step": 12178 }, { "epoch": 1.5492939829538228, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.598773956298828, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.861606240272522, "num_tokens": 464719205.0, "step": 12179 }, { "epoch": 1.5494211932324133, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.53467845916748, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8730892539024353, "num_tokens": 464760758.0, "step": 12180 }, { "epoch": 1.5495484035110036, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.682828903198242, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8721181154251099, "num_tokens": 464797422.0, "step": 12181 }, { "epoch": 1.5496756137895942, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.51069164276123, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8701946139335632, "num_tokens": 464832889.0, "step": 12182 }, { "epoch": 1.5498028240681847, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.642800331115723, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8621311187744141, "num_tokens": 464867743.0, "step": 12183 }, { "epoch": 1.5499300343467752, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.567487716674805, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8559674024581909, "num_tokens": 464904666.0, "step": 12184 }, { "epoch": 1.5500572446253658, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.53977108001709, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8720918893814087, "num_tokens": 464942886.0, "step": 12185 }, { "epoch": 1.5501844549039563, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.688946723937988, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8672758936882019, "num_tokens": 464977463.0, "step": 12186 }, { "epoch": 1.5503116651825466, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.563085556030273, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8777114152908325, "num_tokens": 465015732.0, "step": 12187 }, { "epoch": 1.5504388754611371, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.521081924438477, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8773901462554932, "num_tokens": 465060789.0, "step": 12188 }, { "epoch": 1.5505660857397277, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.666191101074219, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8663644194602966, "num_tokens": 465097907.0, "step": 12189 }, { "epoch": 1.5506932960183182, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.51733684539795, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8771955966949463, "num_tokens": 465139401.0, "step": 12190 }, { "epoch": 1.5508205062969087, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.663291931152344, "learning_rate": 1e-06, "loss": 0.4893, "mean_token_accuracy": 0.8478947281837463, "num_tokens": 465176531.0, "step": 12191 }, { "epoch": 1.5509477165754992, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.600497245788574, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.877206563949585, "num_tokens": 465216074.0, "step": 12192 }, { "epoch": 1.5510749268540898, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.531003952026367, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.881316065788269, "num_tokens": 465257986.0, "step": 12193 }, { "epoch": 1.5512021371326803, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.621320724487305, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8725417852401733, "num_tokens": 465296038.0, "step": 12194 }, { "epoch": 1.5513293474112708, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.581392288208008, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8460243344306946, "num_tokens": 465334215.0, "step": 12195 }, { "epoch": 1.5514565576898613, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.586897850036621, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8677977323532104, "num_tokens": 465373362.0, "step": 12196 }, { "epoch": 1.5515837679684519, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.603028297424316, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8634390234947205, "num_tokens": 465410181.0, "step": 12197 }, { "epoch": 1.5517109782470424, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.650108337402344, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.876712441444397, "num_tokens": 465443056.0, "step": 12198 }, { "epoch": 1.551838188525633, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.556066513061523, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8720136284828186, "num_tokens": 465484509.0, "step": 12199 }, { "epoch": 1.5519653988042235, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.55566120147705, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8674486875534058, "num_tokens": 465520017.0, "step": 12200 }, { "epoch": 1.552092609082814, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.624585151672363, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8849447965621948, "num_tokens": 465557221.0, "step": 12201 }, { "epoch": 1.5522198193614045, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.557650566101074, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8785204887390137, "num_tokens": 465595455.0, "step": 12202 }, { "epoch": 1.552347029639995, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.55794906616211, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8698089718818665, "num_tokens": 465636436.0, "step": 12203 }, { "epoch": 1.5524742399185856, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.625484466552734, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.866161048412323, "num_tokens": 465668956.0, "step": 12204 }, { "epoch": 1.5526014501971759, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.61695384979248, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8496674299240112, "num_tokens": 465708812.0, "step": 12205 }, { "epoch": 1.5527286604757664, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.58458423614502, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8557477593421936, "num_tokens": 465742610.0, "step": 12206 }, { "epoch": 1.552855870754357, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.621320724487305, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8677366375923157, "num_tokens": 465776468.0, "step": 12207 }, { "epoch": 1.5529830810329475, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.59813404083252, "learning_rate": 1e-06, "loss": 0.353, "mean_token_accuracy": 0.8866465091705322, "num_tokens": 465809300.0, "step": 12208 }, { "epoch": 1.553110291311538, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.565258979797363, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8704007863998413, "num_tokens": 465849847.0, "step": 12209 }, { "epoch": 1.5532375015901285, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.580293655395508, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8802391290664673, "num_tokens": 465884986.0, "step": 12210 }, { "epoch": 1.5533647118687188, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.591031074523926, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.876270592212677, "num_tokens": 465919710.0, "step": 12211 }, { "epoch": 1.5534919221473094, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.608840942382812, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.861402690410614, "num_tokens": 465956041.0, "step": 12212 }, { "epoch": 1.5536191324258999, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.502457618713379, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8805491924285889, "num_tokens": 465994352.0, "step": 12213 }, { "epoch": 1.5537463427044904, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.650650024414062, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8689122200012207, "num_tokens": 466032493.0, "step": 12214 }, { "epoch": 1.553873552983081, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.56299877166748, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.871855616569519, "num_tokens": 466067048.0, "step": 12215 }, { "epoch": 1.5540007632616715, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.581502914428711, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8473024964332581, "num_tokens": 466106895.0, "step": 12216 }, { "epoch": 1.554127973540262, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.609943389892578, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8543576598167419, "num_tokens": 466146740.0, "step": 12217 }, { "epoch": 1.5542551838188525, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.586155891418457, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8748936653137207, "num_tokens": 466187097.0, "step": 12218 }, { "epoch": 1.554382394097443, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.559928894042969, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8600583076477051, "num_tokens": 466223969.0, "step": 12219 }, { "epoch": 1.5545096043760336, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.586793899536133, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8799854516983032, "num_tokens": 466260127.0, "step": 12220 }, { "epoch": 1.554636814654624, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.550638198852539, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8748906850814819, "num_tokens": 466297219.0, "step": 12221 }, { "epoch": 1.5547640249332146, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.556937217712402, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.850464403629303, "num_tokens": 466329474.0, "step": 12222 }, { "epoch": 1.5548912352118052, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.586211204528809, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8674759268760681, "num_tokens": 466372801.0, "step": 12223 }, { "epoch": 1.5550184454903957, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.539453506469727, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8754827976226807, "num_tokens": 466412950.0, "step": 12224 }, { "epoch": 1.5551456557689862, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.593804359436035, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8756798505783081, "num_tokens": 466448830.0, "step": 12225 }, { "epoch": 1.5552728660475768, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.568894386291504, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8586443662643433, "num_tokens": 466489749.0, "step": 12226 }, { "epoch": 1.5554000763261673, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.63614273071289, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8509915471076965, "num_tokens": 466531320.0, "step": 12227 }, { "epoch": 1.5555272866047578, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.504536628723145, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8606187701225281, "num_tokens": 466570001.0, "step": 12228 }, { "epoch": 1.5556544968833483, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.587666511535645, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8715558648109436, "num_tokens": 466607186.0, "step": 12229 }, { "epoch": 1.5557817071619386, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.5354585647583, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8600059747695923, "num_tokens": 466649480.0, "step": 12230 }, { "epoch": 1.5559089174405292, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.57595443725586, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8753640651702881, "num_tokens": 466688861.0, "step": 12231 }, { "epoch": 1.5560361277191197, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.607311248779297, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8498135805130005, "num_tokens": 466723200.0, "step": 12232 }, { "epoch": 1.5561633379977102, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.512137413024902, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8545483350753784, "num_tokens": 466759639.0, "step": 12233 }, { "epoch": 1.5562905482763008, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.595970153808594, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8629879951477051, "num_tokens": 466795748.0, "step": 12234 }, { "epoch": 1.5564177585548913, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.605677604675293, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8790571689605713, "num_tokens": 466832938.0, "step": 12235 }, { "epoch": 1.5565449688334816, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.589600563049316, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.862463116645813, "num_tokens": 466868710.0, "step": 12236 }, { "epoch": 1.5566721791120721, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.64229679107666, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8660229444503784, "num_tokens": 466903558.0, "step": 12237 }, { "epoch": 1.5567993893906626, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.539508819580078, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8592692613601685, "num_tokens": 466946785.0, "step": 12238 }, { "epoch": 1.5569265996692532, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.581300735473633, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8608785271644592, "num_tokens": 466988563.0, "step": 12239 }, { "epoch": 1.5570538099478437, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.572611808776855, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.862328052520752, "num_tokens": 467025447.0, "step": 12240 }, { "epoch": 1.5571810202264342, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.563804626464844, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8763642311096191, "num_tokens": 467060751.0, "step": 12241 }, { "epoch": 1.5573082305050248, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.546736717224121, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.867489755153656, "num_tokens": 467101889.0, "step": 12242 }, { "epoch": 1.5574354407836153, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.613349914550781, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8705641031265259, "num_tokens": 467136782.0, "step": 12243 }, { "epoch": 1.5575626510622058, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.598544120788574, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8778810501098633, "num_tokens": 467174137.0, "step": 12244 }, { "epoch": 1.5576898613407963, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.54702377319336, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8645390272140503, "num_tokens": 467213919.0, "step": 12245 }, { "epoch": 1.5578170716193869, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.579434394836426, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8759762048721313, "num_tokens": 467254860.0, "step": 12246 }, { "epoch": 1.5579442818979774, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.583959579467773, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.879024088382721, "num_tokens": 467296806.0, "step": 12247 }, { "epoch": 1.558071492176568, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.570086479187012, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8744106888771057, "num_tokens": 467332882.0, "step": 12248 }, { "epoch": 1.5581987024551585, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53459644317627, "learning_rate": 1e-06, "loss": 0.3595, "mean_token_accuracy": 0.8831989765167236, "num_tokens": 467375088.0, "step": 12249 }, { "epoch": 1.558325912733749, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.599691390991211, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8686665892601013, "num_tokens": 467413000.0, "step": 12250 }, { "epoch": 1.5584531230123395, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.560946464538574, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8827854990959167, "num_tokens": 467450947.0, "step": 12251 }, { "epoch": 1.55858033329093, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.605180740356445, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8585184216499329, "num_tokens": 467493770.0, "step": 12252 }, { "epoch": 1.5587075435695206, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.545001029968262, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8659720420837402, "num_tokens": 467531882.0, "step": 12253 }, { "epoch": 1.5588347538481109, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.618674278259277, "learning_rate": 1e-06, "loss": 0.3187, "mean_token_accuracy": 0.8946685194969177, "num_tokens": 467566059.0, "step": 12254 }, { "epoch": 1.5589619641267014, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.508455276489258, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8832858800888062, "num_tokens": 467603734.0, "step": 12255 }, { "epoch": 1.559089174405292, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.616065979003906, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8762158155441284, "num_tokens": 467641224.0, "step": 12256 }, { "epoch": 1.5592163846838825, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.616145133972168, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8678187727928162, "num_tokens": 467677576.0, "step": 12257 }, { "epoch": 1.559343594962473, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.578661918640137, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8752710223197937, "num_tokens": 467717526.0, "step": 12258 }, { "epoch": 1.5594708052410635, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.541189193725586, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8649132251739502, "num_tokens": 467757591.0, "step": 12259 }, { "epoch": 1.5595980155196538, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.660296440124512, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8549391627311707, "num_tokens": 467792794.0, "step": 12260 }, { "epoch": 1.5597252257982444, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.616666793823242, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8632602095603943, "num_tokens": 467829554.0, "step": 12261 }, { "epoch": 1.5598524360768349, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.54438591003418, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8525916934013367, "num_tokens": 467871403.0, "step": 12262 }, { "epoch": 1.5599796463554254, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.569159507751465, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8491595983505249, "num_tokens": 467906498.0, "step": 12263 }, { "epoch": 1.560106856634016, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.57811450958252, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8669372797012329, "num_tokens": 467946426.0, "step": 12264 }, { "epoch": 1.5602340669126065, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.612286567687988, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8709428310394287, "num_tokens": 467979930.0, "step": 12265 }, { "epoch": 1.560361277191197, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.612669944763184, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8620142340660095, "num_tokens": 468016347.0, "step": 12266 }, { "epoch": 1.5604884874697875, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.525227546691895, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8758813142776489, "num_tokens": 468051446.0, "step": 12267 }, { "epoch": 1.560615697748378, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.580676078796387, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8838372826576233, "num_tokens": 468085713.0, "step": 12268 }, { "epoch": 1.5607429080269686, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.525541305541992, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8809673190116882, "num_tokens": 468117863.0, "step": 12269 }, { "epoch": 1.560870118305559, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.5946626663208, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8651425838470459, "num_tokens": 468153041.0, "step": 12270 }, { "epoch": 1.5609973285841496, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.586915016174316, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.866639256477356, "num_tokens": 468193624.0, "step": 12271 }, { "epoch": 1.5611245388627402, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.605263710021973, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8841422200202942, "num_tokens": 468227091.0, "step": 12272 }, { "epoch": 1.5612517491413307, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.598322868347168, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8636246919631958, "num_tokens": 468265079.0, "step": 12273 }, { "epoch": 1.5613789594199212, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.604269981384277, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8654877543449402, "num_tokens": 468307055.0, "step": 12274 }, { "epoch": 1.5615061696985117, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.625898361206055, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8625057935714722, "num_tokens": 468344429.0, "step": 12275 }, { "epoch": 1.5616333799771023, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.606529235839844, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8682985305786133, "num_tokens": 468379448.0, "step": 12276 }, { "epoch": 1.5617605902556928, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.600850105285645, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8711012005805969, "num_tokens": 468423344.0, "step": 12277 }, { "epoch": 1.5618878005342833, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.563432693481445, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.87385094165802, "num_tokens": 468461657.0, "step": 12278 }, { "epoch": 1.5620150108128736, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.634115219116211, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8832598328590393, "num_tokens": 468500428.0, "step": 12279 }, { "epoch": 1.5621422210914642, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.620725631713867, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8570858240127563, "num_tokens": 468536310.0, "step": 12280 }, { "epoch": 1.5622694313700547, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.616122245788574, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8663763999938965, "num_tokens": 468577954.0, "step": 12281 }, { "epoch": 1.5623966416486452, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.636283874511719, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8598949313163757, "num_tokens": 468612131.0, "step": 12282 }, { "epoch": 1.5625238519272358, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.625618934631348, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8638532757759094, "num_tokens": 468654022.0, "step": 12283 }, { "epoch": 1.5626510622058263, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53079605102539, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.859740138053894, "num_tokens": 468693120.0, "step": 12284 }, { "epoch": 1.5627782724844166, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.64280891418457, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8678079843521118, "num_tokens": 468727695.0, "step": 12285 }, { "epoch": 1.5629054827630071, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.595252990722656, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8657571077346802, "num_tokens": 468766705.0, "step": 12286 }, { "epoch": 1.5630326930415976, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.681842803955078, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8601361513137817, "num_tokens": 468804555.0, "step": 12287 }, { "epoch": 1.5631599033201882, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.588781356811523, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8654583692550659, "num_tokens": 468842495.0, "step": 12288 }, { "epoch": 1.5632871135987787, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.62049674987793, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8674473762512207, "num_tokens": 468877574.0, "step": 12289 }, { "epoch": 1.5634143238773692, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.58969497680664, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8796336054801941, "num_tokens": 468914573.0, "step": 12290 }, { "epoch": 1.5635415341559598, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.615017890930176, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8792625665664673, "num_tokens": 468950943.0, "step": 12291 }, { "epoch": 1.5636687444345503, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.593269348144531, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8832454681396484, "num_tokens": 468990007.0, "step": 12292 }, { "epoch": 1.5637959547131408, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.547805786132812, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8723001480102539, "num_tokens": 469022461.0, "step": 12293 }, { "epoch": 1.5639231649917313, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.646728515625, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8738210201263428, "num_tokens": 469056396.0, "step": 12294 }, { "epoch": 1.5640503752703219, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.604347229003906, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8775256872177124, "num_tokens": 469099464.0, "step": 12295 }, { "epoch": 1.5641775855489124, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.605856895446777, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8775372505187988, "num_tokens": 469139830.0, "step": 12296 }, { "epoch": 1.564304795827503, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.553718566894531, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8677092790603638, "num_tokens": 469177838.0, "step": 12297 }, { "epoch": 1.5644320061060935, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.64545726776123, "learning_rate": 1e-06, "loss": 0.334, "mean_token_accuracy": 0.8933597803115845, "num_tokens": 469218727.0, "step": 12298 }, { "epoch": 1.564559216384684, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.529170989990234, "learning_rate": 1e-06, "loss": 0.4816, "mean_token_accuracy": 0.846302330493927, "num_tokens": 469261186.0, "step": 12299 }, { "epoch": 1.5646864266632745, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.621702194213867, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8765573501586914, "num_tokens": 469301191.0, "step": 12300 }, { "epoch": 1.564813636941865, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.63302993774414, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8739862442016602, "num_tokens": 469340132.0, "step": 12301 }, { "epoch": 1.5649408472204556, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.602312088012695, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.8873194456100464, "num_tokens": 469376211.0, "step": 12302 }, { "epoch": 1.5650680574990459, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.67683219909668, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8587818145751953, "num_tokens": 469413961.0, "step": 12303 }, { "epoch": 1.5651952677776364, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.617830276489258, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8689498901367188, "num_tokens": 469449507.0, "step": 12304 }, { "epoch": 1.565322478056227, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.643464088439941, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8559029698371887, "num_tokens": 469486324.0, "step": 12305 }, { "epoch": 1.5654496883348175, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.6611967086792, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8642300963401794, "num_tokens": 469523799.0, "step": 12306 }, { "epoch": 1.565576898613408, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.62667465209961, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8554078340530396, "num_tokens": 469558661.0, "step": 12307 }, { "epoch": 1.5657041088919985, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.676480293273926, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8673832416534424, "num_tokens": 469603087.0, "step": 12308 }, { "epoch": 1.5658313191705888, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.637836456298828, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8680756092071533, "num_tokens": 469637344.0, "step": 12309 }, { "epoch": 1.5659585294491793, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.527997970581055, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8778985142707825, "num_tokens": 469681127.0, "step": 12310 }, { "epoch": 1.5660857397277699, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.59350299835205, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8675537705421448, "num_tokens": 469724238.0, "step": 12311 }, { "epoch": 1.5662129500063604, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.541271209716797, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8617229461669922, "num_tokens": 469765967.0, "step": 12312 }, { "epoch": 1.566340160284951, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.633744239807129, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8697092533111572, "num_tokens": 469801387.0, "step": 12313 }, { "epoch": 1.5664673705635415, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.594274520874023, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8699055910110474, "num_tokens": 469835448.0, "step": 12314 }, { "epoch": 1.566594580842132, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.591883659362793, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8597052693367004, "num_tokens": 469876647.0, "step": 12315 }, { "epoch": 1.5667217911207225, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.60308837890625, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8828628659248352, "num_tokens": 469918158.0, "step": 12316 }, { "epoch": 1.566849001399313, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.544424057006836, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8549879789352417, "num_tokens": 469953691.0, "step": 12317 }, { "epoch": 1.5669762116779036, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.635343551635742, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8711697459220886, "num_tokens": 469996174.0, "step": 12318 }, { "epoch": 1.567103421956494, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.585976600646973, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8558336496353149, "num_tokens": 470031371.0, "step": 12319 }, { "epoch": 1.5672306322350846, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.650877952575684, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8725670576095581, "num_tokens": 470068783.0, "step": 12320 }, { "epoch": 1.5673578425136752, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.634150505065918, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.851539134979248, "num_tokens": 470103490.0, "step": 12321 }, { "epoch": 1.5674850527922657, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.510937690734863, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8650786876678467, "num_tokens": 470137213.0, "step": 12322 }, { "epoch": 1.5676122630708562, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.614654541015625, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8726006150245667, "num_tokens": 470171816.0, "step": 12323 }, { "epoch": 1.5677394733494467, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.582509994506836, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8460111618041992, "num_tokens": 470210517.0, "step": 12324 }, { "epoch": 1.5678666836280373, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.61739730834961, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8563119769096375, "num_tokens": 470247085.0, "step": 12325 }, { "epoch": 1.5679938939066278, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.61977767944336, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.868393063545227, "num_tokens": 470288483.0, "step": 12326 }, { "epoch": 1.5681211041852183, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.626869201660156, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8676414489746094, "num_tokens": 470322173.0, "step": 12327 }, { "epoch": 1.5682483144638086, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.616106033325195, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8699483871459961, "num_tokens": 470356617.0, "step": 12328 }, { "epoch": 1.5683755247423992, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.568790435791016, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8799230456352234, "num_tokens": 470393095.0, "step": 12329 }, { "epoch": 1.5685027350209897, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.651728630065918, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8797827959060669, "num_tokens": 470427785.0, "step": 12330 }, { "epoch": 1.5686299452995802, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.598138809204102, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8632951974868774, "num_tokens": 470469971.0, "step": 12331 }, { "epoch": 1.5687571555781707, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.703827857971191, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8608571290969849, "num_tokens": 470503935.0, "step": 12332 }, { "epoch": 1.5688843658567613, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.58714771270752, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.871360719203949, "num_tokens": 470536014.0, "step": 12333 }, { "epoch": 1.5690115761353516, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.55799674987793, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8810458183288574, "num_tokens": 470575632.0, "step": 12334 }, { "epoch": 1.569138786413942, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.595905303955078, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8793820142745972, "num_tokens": 470611295.0, "step": 12335 }, { "epoch": 1.5692659966925326, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.615224838256836, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.867135763168335, "num_tokens": 470646993.0, "step": 12336 }, { "epoch": 1.5693932069711232, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.594524383544922, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8726126551628113, "num_tokens": 470680253.0, "step": 12337 }, { "epoch": 1.5695204172497137, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.610243797302246, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8665252327919006, "num_tokens": 470716169.0, "step": 12338 }, { "epoch": 1.5696476275283042, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.619112014770508, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8810839653015137, "num_tokens": 470754208.0, "step": 12339 }, { "epoch": 1.5697748378068948, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.613030433654785, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8703169822692871, "num_tokens": 470801329.0, "step": 12340 }, { "epoch": 1.5699020480854853, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.644185066223145, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8643721342086792, "num_tokens": 470838408.0, "step": 12341 }, { "epoch": 1.5700292583640758, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.616130828857422, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8690602779388428, "num_tokens": 470874001.0, "step": 12342 }, { "epoch": 1.5701564686426663, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.611170768737793, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8638355731964111, "num_tokens": 470916870.0, "step": 12343 }, { "epoch": 1.5702836789212569, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.628679275512695, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8769479393959045, "num_tokens": 470953862.0, "step": 12344 }, { "epoch": 1.5704108891998474, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.58796215057373, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8788397312164307, "num_tokens": 470991507.0, "step": 12345 }, { "epoch": 1.570538099478438, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.658451080322266, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8719719648361206, "num_tokens": 471030109.0, "step": 12346 }, { "epoch": 1.5706653097570284, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.626912117004395, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8816416263580322, "num_tokens": 471063829.0, "step": 12347 }, { "epoch": 1.570792520035619, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.670985221862793, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8585378527641296, "num_tokens": 471107259.0, "step": 12348 }, { "epoch": 1.5709197303142095, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.669570922851562, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8721469640731812, "num_tokens": 471141516.0, "step": 12349 }, { "epoch": 1.5710469405928, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.665620803833008, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8697320818901062, "num_tokens": 471178655.0, "step": 12350 }, { "epoch": 1.5711741508713906, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.682801246643066, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8799611926078796, "num_tokens": 471209437.0, "step": 12351 }, { "epoch": 1.5713013611499809, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.522175788879395, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8839030861854553, "num_tokens": 471249559.0, "step": 12352 }, { "epoch": 1.5714285714285714, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.70579719543457, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8742588758468628, "num_tokens": 471287179.0, "step": 12353 }, { "epoch": 1.571555781707162, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.605913162231445, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8727261424064636, "num_tokens": 471328052.0, "step": 12354 }, { "epoch": 1.5716829919857525, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.615225791931152, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8669717311859131, "num_tokens": 471362414.0, "step": 12355 }, { "epoch": 1.571810202264343, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.636458396911621, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8656572103500366, "num_tokens": 471397910.0, "step": 12356 }, { "epoch": 1.5719374125429335, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.67209243774414, "learning_rate": 1e-06, "loss": 0.3601, "mean_token_accuracy": 0.8833149671554565, "num_tokens": 471436951.0, "step": 12357 }, { "epoch": 1.5720646228215238, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.603264808654785, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8720525503158569, "num_tokens": 471470139.0, "step": 12358 }, { "epoch": 1.5721918331001143, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.657381057739258, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8769552111625671, "num_tokens": 471508591.0, "step": 12359 }, { "epoch": 1.5723190433787049, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.588173866271973, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8702491521835327, "num_tokens": 471549165.0, "step": 12360 }, { "epoch": 1.5724462536572954, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.581461906433105, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8642557859420776, "num_tokens": 471589021.0, "step": 12361 }, { "epoch": 1.572573463935886, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.691535949707031, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8636201620101929, "num_tokens": 471624263.0, "step": 12362 }, { "epoch": 1.5727006742144765, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.571953773498535, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8641176819801331, "num_tokens": 471663678.0, "step": 12363 }, { "epoch": 1.572827884493067, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.592975616455078, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8674591183662415, "num_tokens": 471705887.0, "step": 12364 }, { "epoch": 1.5729550947716575, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.656588554382324, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8774017095565796, "num_tokens": 471739796.0, "step": 12365 }, { "epoch": 1.573082305050248, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.634187698364258, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8661232590675354, "num_tokens": 471774723.0, "step": 12366 }, { "epoch": 1.5732095153288386, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.619752883911133, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8623338937759399, "num_tokens": 471814169.0, "step": 12367 }, { "epoch": 1.573336725607429, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.61356258392334, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8701151609420776, "num_tokens": 471852367.0, "step": 12368 }, { "epoch": 1.5734639358860196, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.61807918548584, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8746120929718018, "num_tokens": 471883438.0, "step": 12369 }, { "epoch": 1.5735911461646102, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.584693908691406, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8650137186050415, "num_tokens": 471921051.0, "step": 12370 }, { "epoch": 1.5737183564432007, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.663542747497559, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8735306262969971, "num_tokens": 471959685.0, "step": 12371 }, { "epoch": 1.5738455667217912, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.591561317443848, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8700390458106995, "num_tokens": 471997345.0, "step": 12372 }, { "epoch": 1.5739727770003817, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.662582397460938, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.874793529510498, "num_tokens": 472034455.0, "step": 12373 }, { "epoch": 1.5740999872789723, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.636212348937988, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8694977760314941, "num_tokens": 472073630.0, "step": 12374 }, { "epoch": 1.5742271975575628, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.65225601196289, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8676724433898926, "num_tokens": 472111556.0, "step": 12375 }, { "epoch": 1.5743544078361533, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.597707748413086, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8765364289283752, "num_tokens": 472148276.0, "step": 12376 }, { "epoch": 1.5744816181147436, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.588356971740723, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.869276762008667, "num_tokens": 472181665.0, "step": 12377 }, { "epoch": 1.5746088283933342, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.595309257507324, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8635898232460022, "num_tokens": 472222759.0, "step": 12378 }, { "epoch": 1.5747360386719247, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.601921081542969, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8635788559913635, "num_tokens": 472260867.0, "step": 12379 }, { "epoch": 1.5748632489505152, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.633955955505371, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8577375411987305, "num_tokens": 472300068.0, "step": 12380 }, { "epoch": 1.5749904592291057, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.590739250183105, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8808003664016724, "num_tokens": 472336459.0, "step": 12381 }, { "epoch": 1.5751176695076963, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.567411422729492, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8738139271736145, "num_tokens": 472376696.0, "step": 12382 }, { "epoch": 1.5752448797862866, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.614303588867188, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8822057247161865, "num_tokens": 472414491.0, "step": 12383 }, { "epoch": 1.575372090064877, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.646310806274414, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8691349029541016, "num_tokens": 472452188.0, "step": 12384 }, { "epoch": 1.5754993003434676, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.603784561157227, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8678908348083496, "num_tokens": 472489690.0, "step": 12385 }, { "epoch": 1.5756265106220582, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.594278335571289, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.868137538433075, "num_tokens": 472529912.0, "step": 12386 }, { "epoch": 1.5757537209006487, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.648490905761719, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8680444955825806, "num_tokens": 472562534.0, "step": 12387 }, { "epoch": 1.5758809311792392, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.679839134216309, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8676836490631104, "num_tokens": 472598955.0, "step": 12388 }, { "epoch": 1.5760081414578297, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.593169212341309, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8734413385391235, "num_tokens": 472638064.0, "step": 12389 }, { "epoch": 1.5761353517364203, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.600018501281738, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8734117150306702, "num_tokens": 472676823.0, "step": 12390 }, { "epoch": 1.5762625620150108, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.644209861755371, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8694319725036621, "num_tokens": 472718528.0, "step": 12391 }, { "epoch": 1.5763897722936013, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.613401412963867, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8755122423171997, "num_tokens": 472755634.0, "step": 12392 }, { "epoch": 1.5765169825721919, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.63125991821289, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8804677128791809, "num_tokens": 472798060.0, "step": 12393 }, { "epoch": 1.5766441928507824, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.524796485900879, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8751220703125, "num_tokens": 472837628.0, "step": 12394 }, { "epoch": 1.576771403129373, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.684835433959961, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8705384135246277, "num_tokens": 472877446.0, "step": 12395 }, { "epoch": 1.5768986134079634, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.541492462158203, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8639564514160156, "num_tokens": 472918437.0, "step": 12396 }, { "epoch": 1.577025823686554, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.628095626831055, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8779107332229614, "num_tokens": 472956242.0, "step": 12397 }, { "epoch": 1.5771530339651445, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.710797309875488, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8787400722503662, "num_tokens": 472985620.0, "step": 12398 }, { "epoch": 1.577280244243735, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.586647987365723, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8743338584899902, "num_tokens": 473019517.0, "step": 12399 }, { "epoch": 1.5774074545223256, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.541420936584473, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.8885955810546875, "num_tokens": 473054624.0, "step": 12400 }, { "epoch": 1.5775346648009159, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.623775482177734, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8601922988891602, "num_tokens": 473094140.0, "step": 12401 }, { "epoch": 1.5776618750795064, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.58792495727539, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8511396646499634, "num_tokens": 473136886.0, "step": 12402 }, { "epoch": 1.577789085358097, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.614161491394043, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8580745458602905, "num_tokens": 473173267.0, "step": 12403 }, { "epoch": 1.5779162956366874, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.616140365600586, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8508466482162476, "num_tokens": 473216477.0, "step": 12404 }, { "epoch": 1.578043505915278, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.586177825927734, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8618909120559692, "num_tokens": 473256986.0, "step": 12405 }, { "epoch": 1.5781707161938685, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.666787147521973, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8776809573173523, "num_tokens": 473297967.0, "step": 12406 }, { "epoch": 1.5782979264724588, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.611733436584473, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8620375394821167, "num_tokens": 473333580.0, "step": 12407 }, { "epoch": 1.5784251367510493, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.617456436157227, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8744164705276489, "num_tokens": 473370553.0, "step": 12408 }, { "epoch": 1.5785523470296399, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.6442289352417, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8520938754081726, "num_tokens": 473410705.0, "step": 12409 }, { "epoch": 1.5786795573082304, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.580873489379883, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8744032382965088, "num_tokens": 473443315.0, "step": 12410 }, { "epoch": 1.578806767586821, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.593646049499512, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8613809943199158, "num_tokens": 473488767.0, "step": 12411 }, { "epoch": 1.5789339778654115, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.599512100219727, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8845770955085754, "num_tokens": 473525360.0, "step": 12412 }, { "epoch": 1.579061188144002, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.64671802520752, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.862009584903717, "num_tokens": 473560625.0, "step": 12413 }, { "epoch": 1.5791883984225925, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.52277660369873, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8887784481048584, "num_tokens": 473595484.0, "step": 12414 }, { "epoch": 1.579315608701183, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.643742561340332, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8673086166381836, "num_tokens": 473627269.0, "step": 12415 }, { "epoch": 1.5794428189797736, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.678913116455078, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8545277714729309, "num_tokens": 473669159.0, "step": 12416 }, { "epoch": 1.579570029258364, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.615130424499512, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8793643116950989, "num_tokens": 473707739.0, "step": 12417 }, { "epoch": 1.5796972395369546, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.58447551727295, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8708523511886597, "num_tokens": 473746908.0, "step": 12418 }, { "epoch": 1.5798244498155452, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.636340141296387, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8677743077278137, "num_tokens": 473781480.0, "step": 12419 }, { "epoch": 1.5799516600941357, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.63911247253418, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8684555292129517, "num_tokens": 473814676.0, "step": 12420 }, { "epoch": 1.5800788703727262, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.628939628601074, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8649968504905701, "num_tokens": 473855487.0, "step": 12421 }, { "epoch": 1.5802060806513167, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.644124984741211, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8812676072120667, "num_tokens": 473885648.0, "step": 12422 }, { "epoch": 1.5803332909299073, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.612602233886719, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8646899461746216, "num_tokens": 473923862.0, "step": 12423 }, { "epoch": 1.5804605012084978, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.607866287231445, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8731955289840698, "num_tokens": 473957725.0, "step": 12424 }, { "epoch": 1.5805877114870883, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.584356307983398, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8711158037185669, "num_tokens": 473994939.0, "step": 12425 }, { "epoch": 1.5807149217656786, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.612661361694336, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8658857345581055, "num_tokens": 474035996.0, "step": 12426 }, { "epoch": 1.5808421320442692, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.594279289245605, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8783976435661316, "num_tokens": 474071693.0, "step": 12427 }, { "epoch": 1.5809693423228597, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.59490966796875, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8755242824554443, "num_tokens": 474114415.0, "step": 12428 }, { "epoch": 1.5810965526014502, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.651619911193848, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8640660047531128, "num_tokens": 474152060.0, "step": 12429 }, { "epoch": 1.5812237628800407, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.56910228729248, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8735818862915039, "num_tokens": 474190605.0, "step": 12430 }, { "epoch": 1.5813509731586313, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.607791900634766, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8651973009109497, "num_tokens": 474229447.0, "step": 12431 }, { "epoch": 1.5814781834372216, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.61331558227539, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8780655860900879, "num_tokens": 474268088.0, "step": 12432 }, { "epoch": 1.581605393715812, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.580254554748535, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8655298948287964, "num_tokens": 474304856.0, "step": 12433 }, { "epoch": 1.5817326039944026, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.60861873626709, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.872174859046936, "num_tokens": 474349248.0, "step": 12434 }, { "epoch": 1.5818598142729932, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.571317672729492, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8802600502967834, "num_tokens": 474384432.0, "step": 12435 }, { "epoch": 1.5819870245515837, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.59482479095459, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8675501346588135, "num_tokens": 474415881.0, "step": 12436 }, { "epoch": 1.5821142348301742, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.65817928314209, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8677408695220947, "num_tokens": 474452755.0, "step": 12437 }, { "epoch": 1.5822414451087647, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.557599067687988, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8654466867446899, "num_tokens": 474493937.0, "step": 12438 }, { "epoch": 1.5823686553873553, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.673637390136719, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8758818507194519, "num_tokens": 474534072.0, "step": 12439 }, { "epoch": 1.5824958656659458, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.612706184387207, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8625698685646057, "num_tokens": 474569343.0, "step": 12440 }, { "epoch": 1.5826230759445363, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.588399887084961, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8595472574234009, "num_tokens": 474606610.0, "step": 12441 }, { "epoch": 1.5827502862231269, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.655372619628906, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.861346960067749, "num_tokens": 474645071.0, "step": 12442 }, { "epoch": 1.5828774965017174, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.619534492492676, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8568726181983948, "num_tokens": 474683985.0, "step": 12443 }, { "epoch": 1.583004706780308, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.588567733764648, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8757354617118835, "num_tokens": 474726102.0, "step": 12444 }, { "epoch": 1.5831319170588984, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.628185272216797, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8718461990356445, "num_tokens": 474766430.0, "step": 12445 }, { "epoch": 1.583259127337489, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.62511157989502, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8692053556442261, "num_tokens": 474800648.0, "step": 12446 }, { "epoch": 1.5833863376160795, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.623442649841309, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8732022643089294, "num_tokens": 474845540.0, "step": 12447 }, { "epoch": 1.58351354789467, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.593774795532227, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8673396110534668, "num_tokens": 474884296.0, "step": 12448 }, { "epoch": 1.5836407581732606, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.637784957885742, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8739802837371826, "num_tokens": 474929344.0, "step": 12449 }, { "epoch": 1.5837679684518509, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.704547882080078, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8750620484352112, "num_tokens": 474968035.0, "step": 12450 }, { "epoch": 1.5838951787304414, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.672789573669434, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8639163374900818, "num_tokens": 475006189.0, "step": 12451 }, { "epoch": 1.584022389009032, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.598675727844238, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8640474677085876, "num_tokens": 475039768.0, "step": 12452 }, { "epoch": 1.5841495992876224, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.650790214538574, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8745604157447815, "num_tokens": 475079938.0, "step": 12453 }, { "epoch": 1.584276809566213, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.64979362487793, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8574956655502319, "num_tokens": 475122405.0, "step": 12454 }, { "epoch": 1.5844040198448035, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.723898887634277, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8623906373977661, "num_tokens": 475158237.0, "step": 12455 }, { "epoch": 1.5845312301233938, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.610224723815918, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8703116178512573, "num_tokens": 475194966.0, "step": 12456 }, { "epoch": 1.5846584404019843, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.700505256652832, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8685199618339539, "num_tokens": 475234483.0, "step": 12457 }, { "epoch": 1.5847856506805749, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.556331634521484, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8646820783615112, "num_tokens": 475275172.0, "step": 12458 }, { "epoch": 1.5849128609591654, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.721578598022461, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8739334344863892, "num_tokens": 475313550.0, "step": 12459 }, { "epoch": 1.585040071237756, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.621849060058594, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8761001229286194, "num_tokens": 475357019.0, "step": 12460 }, { "epoch": 1.5851672815163464, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.668484687805176, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8737316727638245, "num_tokens": 475394788.0, "step": 12461 }, { "epoch": 1.585294491794937, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.610437393188477, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8675278425216675, "num_tokens": 475432354.0, "step": 12462 }, { "epoch": 1.5854217020735275, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.7700777053833, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8728864192962646, "num_tokens": 475467284.0, "step": 12463 }, { "epoch": 1.585548912352118, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.644572257995605, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8622633218765259, "num_tokens": 475505946.0, "step": 12464 }, { "epoch": 1.5856761226307086, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.653409004211426, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8564436435699463, "num_tokens": 475542799.0, "step": 12465 }, { "epoch": 1.585803332909299, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.651745796203613, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8699132800102234, "num_tokens": 475582064.0, "step": 12466 }, { "epoch": 1.5859305431878896, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.651447296142578, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8650433421134949, "num_tokens": 475615521.0, "step": 12467 }, { "epoch": 1.5860577534664801, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.613154411315918, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8781083822250366, "num_tokens": 475650500.0, "step": 12468 }, { "epoch": 1.5861849637450707, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.629935264587402, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8710391521453857, "num_tokens": 475695824.0, "step": 12469 }, { "epoch": 1.5863121740236612, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.677166938781738, "learning_rate": 1e-06, "loss": 0.4842, "mean_token_accuracy": 0.8474240899085999, "num_tokens": 475734015.0, "step": 12470 }, { "epoch": 1.5864393843022517, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.695548057556152, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8669021129608154, "num_tokens": 475768961.0, "step": 12471 }, { "epoch": 1.5865665945808423, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.57967472076416, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8835440874099731, "num_tokens": 475806551.0, "step": 12472 }, { "epoch": 1.5866938048594328, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.586090087890625, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8622188568115234, "num_tokens": 475844066.0, "step": 12473 }, { "epoch": 1.5868210151380233, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.704412460327148, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8738123178482056, "num_tokens": 475874554.0, "step": 12474 }, { "epoch": 1.5869482254166136, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.598936080932617, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8758018612861633, "num_tokens": 475915840.0, "step": 12475 }, { "epoch": 1.5870754356952042, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.642040252685547, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8675827383995056, "num_tokens": 475951516.0, "step": 12476 }, { "epoch": 1.5872026459737947, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.655817985534668, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.871393084526062, "num_tokens": 475984224.0, "step": 12477 }, { "epoch": 1.5873298562523852, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.671664237976074, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8474419116973877, "num_tokens": 476015758.0, "step": 12478 }, { "epoch": 1.5874570665309757, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.59512710571289, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.866303563117981, "num_tokens": 476051835.0, "step": 12479 }, { "epoch": 1.5875842768095663, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.626520156860352, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8653680682182312, "num_tokens": 476090667.0, "step": 12480 }, { "epoch": 1.5877114870881566, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.6781005859375, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8665982484817505, "num_tokens": 476130448.0, "step": 12481 }, { "epoch": 1.587838697366747, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.64876937866211, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8599943518638611, "num_tokens": 476174495.0, "step": 12482 }, { "epoch": 1.5879659076453376, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.662323951721191, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8776474595069885, "num_tokens": 476213412.0, "step": 12483 }, { "epoch": 1.5880931179239282, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.677722930908203, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8824799656867981, "num_tokens": 476250441.0, "step": 12484 }, { "epoch": 1.5882203282025187, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.555208206176758, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8827693462371826, "num_tokens": 476284466.0, "step": 12485 }, { "epoch": 1.5883475384811092, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.663926124572754, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8673356175422668, "num_tokens": 476327760.0, "step": 12486 }, { "epoch": 1.5884747487596997, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.605535507202148, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8648208975791931, "num_tokens": 476369259.0, "step": 12487 }, { "epoch": 1.5886019590382903, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.647573471069336, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.869752049446106, "num_tokens": 476403915.0, "step": 12488 }, { "epoch": 1.5887291693168808, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.68881893157959, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8675044775009155, "num_tokens": 476440837.0, "step": 12489 }, { "epoch": 1.5888563795954713, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.70977783203125, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8673467636108398, "num_tokens": 476478478.0, "step": 12490 }, { "epoch": 1.5889835898740619, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.667797088623047, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8781168460845947, "num_tokens": 476514675.0, "step": 12491 }, { "epoch": 1.5891108001526524, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.63352108001709, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8674042224884033, "num_tokens": 476559015.0, "step": 12492 }, { "epoch": 1.589238010431243, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.645663261413574, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8546940684318542, "num_tokens": 476601848.0, "step": 12493 }, { "epoch": 1.5893652207098334, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.702547073364258, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8779568672180176, "num_tokens": 476636794.0, "step": 12494 }, { "epoch": 1.589492430988424, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.661245346069336, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.863430917263031, "num_tokens": 476671593.0, "step": 12495 }, { "epoch": 1.5896196412670145, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.65064525604248, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8790189027786255, "num_tokens": 476706776.0, "step": 12496 }, { "epoch": 1.589746851545605, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.683239936828613, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8678243160247803, "num_tokens": 476746700.0, "step": 12497 }, { "epoch": 1.5898740618241956, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.623565673828125, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8744872808456421, "num_tokens": 476785837.0, "step": 12498 }, { "epoch": 1.5900012721027859, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.681756973266602, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8744323253631592, "num_tokens": 476818894.0, "step": 12499 }, { "epoch": 1.5901284823813764, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.627669334411621, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8518131971359253, "num_tokens": 476861773.0, "step": 12500 }, { "epoch": 1.590255692659967, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.698402404785156, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8745250701904297, "num_tokens": 476899522.0, "step": 12501 }, { "epoch": 1.5903829029385574, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.713614463806152, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8625509142875671, "num_tokens": 476939923.0, "step": 12502 }, { "epoch": 1.590510113217148, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.663153648376465, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8682684302330017, "num_tokens": 476979867.0, "step": 12503 }, { "epoch": 1.5906373234957385, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.693197250366211, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.877287745475769, "num_tokens": 477016068.0, "step": 12504 }, { "epoch": 1.5907645337743288, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.668045043945312, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8662869930267334, "num_tokens": 477054526.0, "step": 12505 }, { "epoch": 1.5908917440529193, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.697580337524414, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.864120364189148, "num_tokens": 477098029.0, "step": 12506 }, { "epoch": 1.5910189543315099, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.698047637939453, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8754426836967468, "num_tokens": 477137070.0, "step": 12507 }, { "epoch": 1.5911461646101004, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.729891777038574, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8677608370780945, "num_tokens": 477175640.0, "step": 12508 }, { "epoch": 1.591273374888691, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.695676803588867, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8717328906059265, "num_tokens": 477210935.0, "step": 12509 }, { "epoch": 1.5914005851672814, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.706135749816895, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.875343918800354, "num_tokens": 477251567.0, "step": 12510 }, { "epoch": 1.591527795445872, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.700237274169922, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8720618486404419, "num_tokens": 477287964.0, "step": 12511 }, { "epoch": 1.5916550057244625, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.744853019714355, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8439304828643799, "num_tokens": 477328660.0, "step": 12512 }, { "epoch": 1.591782216003053, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.679610252380371, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8738363981246948, "num_tokens": 477364111.0, "step": 12513 }, { "epoch": 1.5919094262816436, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.648520469665527, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8761601448059082, "num_tokens": 477402691.0, "step": 12514 }, { "epoch": 1.592036636560234, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.710614204406738, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.882393479347229, "num_tokens": 477437526.0, "step": 12515 }, { "epoch": 1.5921638468388246, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.680062294006348, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8676273226737976, "num_tokens": 477479175.0, "step": 12516 }, { "epoch": 1.5922910571174151, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.688440322875977, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.872465968132019, "num_tokens": 477516896.0, "step": 12517 }, { "epoch": 1.5924182673960057, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.689268112182617, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8596668243408203, "num_tokens": 477555543.0, "step": 12518 }, { "epoch": 1.5925454776745962, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.744060516357422, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8807142376899719, "num_tokens": 477592251.0, "step": 12519 }, { "epoch": 1.5926726879531867, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.717222213745117, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8565003871917725, "num_tokens": 477630039.0, "step": 12520 }, { "epoch": 1.5927998982317773, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.768502235412598, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8680533766746521, "num_tokens": 477661028.0, "step": 12521 }, { "epoch": 1.5929271085103678, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.663393020629883, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8824071884155273, "num_tokens": 477699612.0, "step": 12522 }, { "epoch": 1.5930543187889583, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.722718238830566, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8647037148475647, "num_tokens": 477737300.0, "step": 12523 }, { "epoch": 1.5931815290675486, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.70654582977295, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8575016856193542, "num_tokens": 477776031.0, "step": 12524 }, { "epoch": 1.5933087393461391, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.718945503234863, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.870819628238678, "num_tokens": 477813640.0, "step": 12525 }, { "epoch": 1.5934359496247297, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.693397521972656, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.865200400352478, "num_tokens": 477848652.0, "step": 12526 }, { "epoch": 1.5935631599033202, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.720015525817871, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8648754954338074, "num_tokens": 477883710.0, "step": 12527 }, { "epoch": 1.5936903701819107, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.667374610900879, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8628161549568176, "num_tokens": 477920316.0, "step": 12528 }, { "epoch": 1.5938175804605013, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.67728328704834, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8783860802650452, "num_tokens": 477955523.0, "step": 12529 }, { "epoch": 1.5939447907390916, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.720651626586914, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8614077568054199, "num_tokens": 477995768.0, "step": 12530 }, { "epoch": 1.594072001017682, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.65757942199707, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8692550659179688, "num_tokens": 478037718.0, "step": 12531 }, { "epoch": 1.5941992112962726, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.70590877532959, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8580114245414734, "num_tokens": 478077388.0, "step": 12532 }, { "epoch": 1.5943264215748632, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.689809799194336, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8838344812393188, "num_tokens": 478109494.0, "step": 12533 }, { "epoch": 1.5944536318534537, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.716361999511719, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8646234273910522, "num_tokens": 478142420.0, "step": 12534 }, { "epoch": 1.5945808421320442, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.695775032043457, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8739845752716064, "num_tokens": 478183848.0, "step": 12535 }, { "epoch": 1.5947080524106347, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.704781532287598, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8783684968948364, "num_tokens": 478225242.0, "step": 12536 }, { "epoch": 1.5948352626892253, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.697021484375, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8762809038162231, "num_tokens": 478263167.0, "step": 12537 }, { "epoch": 1.5949624729678158, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.720246315002441, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8699698448181152, "num_tokens": 478297752.0, "step": 12538 }, { "epoch": 1.5950896832464063, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.731451988220215, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8700008988380432, "num_tokens": 478332955.0, "step": 12539 }, { "epoch": 1.5952168935249968, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.68594741821289, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8492112755775452, "num_tokens": 478367298.0, "step": 12540 }, { "epoch": 1.5953441038035874, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.731963157653809, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8730594515800476, "num_tokens": 478411183.0, "step": 12541 }, { "epoch": 1.595471314082178, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.708062171936035, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.867700457572937, "num_tokens": 478451184.0, "step": 12542 }, { "epoch": 1.5955985243607684, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.719491958618164, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8646748065948486, "num_tokens": 478490913.0, "step": 12543 }, { "epoch": 1.595725734639359, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.72110366821289, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8705500364303589, "num_tokens": 478529769.0, "step": 12544 }, { "epoch": 1.5958529449179495, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.696992874145508, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8663883209228516, "num_tokens": 478566665.0, "step": 12545 }, { "epoch": 1.59598015519654, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.67429256439209, "learning_rate": 1e-06, "loss": 0.326, "mean_token_accuracy": 0.8951728343963623, "num_tokens": 478600814.0, "step": 12546 }, { "epoch": 1.5961073654751305, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.763398170471191, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8722899556159973, "num_tokens": 478641610.0, "step": 12547 }, { "epoch": 1.5962345757537209, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.664459228515625, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8775228261947632, "num_tokens": 478678183.0, "step": 12548 }, { "epoch": 1.5963617860323114, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.74865436553955, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8703280091285706, "num_tokens": 478710881.0, "step": 12549 }, { "epoch": 1.596488996310902, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.756787300109863, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8837919235229492, "num_tokens": 478748550.0, "step": 12550 }, { "epoch": 1.5966162065894924, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.730329513549805, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8576472997665405, "num_tokens": 478790417.0, "step": 12551 }, { "epoch": 1.596743416868083, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.712919235229492, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8758442401885986, "num_tokens": 478828856.0, "step": 12552 }, { "epoch": 1.5968706271466735, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.751227378845215, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8662252426147461, "num_tokens": 478870240.0, "step": 12553 }, { "epoch": 1.5969978374252638, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.691987037658691, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8616465330123901, "num_tokens": 478911920.0, "step": 12554 }, { "epoch": 1.5971250477038543, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.775422096252441, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8636989593505859, "num_tokens": 478949083.0, "step": 12555 }, { "epoch": 1.5972522579824449, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.787503242492676, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8683657050132751, "num_tokens": 478984087.0, "step": 12556 }, { "epoch": 1.5973794682610354, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.749865531921387, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8503172397613525, "num_tokens": 479019364.0, "step": 12557 }, { "epoch": 1.597506678539626, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.806419372558594, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8728601932525635, "num_tokens": 479059002.0, "step": 12558 }, { "epoch": 1.5976338888182164, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.733257293701172, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8684065341949463, "num_tokens": 479092676.0, "step": 12559 }, { "epoch": 1.597761099096807, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.731350898742676, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8722327351570129, "num_tokens": 479125951.0, "step": 12560 }, { "epoch": 1.5978883093753975, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.744333267211914, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.884497880935669, "num_tokens": 479166230.0, "step": 12561 }, { "epoch": 1.598015519653988, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.65755844116211, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8657928705215454, "num_tokens": 479205813.0, "step": 12562 }, { "epoch": 1.5981427299325786, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.778704643249512, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8656105399131775, "num_tokens": 479244960.0, "step": 12563 }, { "epoch": 1.598269940211169, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.63260269165039, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8758434057235718, "num_tokens": 479289617.0, "step": 12564 }, { "epoch": 1.5983971504897596, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.72146224975586, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8694924116134644, "num_tokens": 479326377.0, "step": 12565 }, { "epoch": 1.5985243607683501, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.73210620880127, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8691356182098389, "num_tokens": 479360563.0, "step": 12566 }, { "epoch": 1.5986515710469407, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.619930267333984, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8810082674026489, "num_tokens": 479403042.0, "step": 12567 }, { "epoch": 1.5987787813255312, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.70640754699707, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8745917081832886, "num_tokens": 479437099.0, "step": 12568 }, { "epoch": 1.5989059916041217, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.743674278259277, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8686748147010803, "num_tokens": 479479326.0, "step": 12569 }, { "epoch": 1.5990332018827123, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.616764068603516, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8525524735450745, "num_tokens": 479515320.0, "step": 12570 }, { "epoch": 1.5991604121613028, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.711883544921875, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8581363558769226, "num_tokens": 479559960.0, "step": 12571 }, { "epoch": 1.5992876224398933, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.575652122497559, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8700233697891235, "num_tokens": 479599115.0, "step": 12572 }, { "epoch": 1.5994148327184836, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.705826759338379, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8772940039634705, "num_tokens": 479636116.0, "step": 12573 }, { "epoch": 1.5995420429970741, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.69951057434082, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8613418936729431, "num_tokens": 479676880.0, "step": 12574 }, { "epoch": 1.5996692532756647, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.670675277709961, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8713177442550659, "num_tokens": 479717424.0, "step": 12575 }, { "epoch": 1.5997964635542552, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.69104290008545, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8678319454193115, "num_tokens": 479756240.0, "step": 12576 }, { "epoch": 1.5999236738328457, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.682022094726562, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8544083833694458, "num_tokens": 479791863.0, "step": 12577 }, { "epoch": 1.6000508841114363, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.632530212402344, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8793658018112183, "num_tokens": 479827252.0, "step": 12578 }, { "epoch": 1.6001780943900266, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.723942756652832, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8659873008728027, "num_tokens": 479863100.0, "step": 12579 }, { "epoch": 1.600305304668617, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.655441284179688, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8590955138206482, "num_tokens": 479899684.0, "step": 12580 }, { "epoch": 1.6004325149472076, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.755513191223145, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8702661991119385, "num_tokens": 479937656.0, "step": 12581 }, { "epoch": 1.6005597252257981, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.750372886657715, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8647278547286987, "num_tokens": 479973357.0, "step": 12582 }, { "epoch": 1.6006869355043887, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.698907852172852, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8704203367233276, "num_tokens": 480017862.0, "step": 12583 }, { "epoch": 1.6008141457829792, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.746716499328613, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8759186267852783, "num_tokens": 480057039.0, "step": 12584 }, { "epoch": 1.6009413560615697, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.704517364501953, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8723275065422058, "num_tokens": 480093396.0, "step": 12585 }, { "epoch": 1.6010685663401603, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.80153751373291, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8656045198440552, "num_tokens": 480125200.0, "step": 12586 }, { "epoch": 1.6011957766187508, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.7298002243042, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8670505881309509, "num_tokens": 480158085.0, "step": 12587 }, { "epoch": 1.6013229868973413, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.807769775390625, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.884390652179718, "num_tokens": 480192979.0, "step": 12588 }, { "epoch": 1.6014501971759318, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.69546127319336, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8794006109237671, "num_tokens": 480227984.0, "step": 12589 }, { "epoch": 1.6015774074545224, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.755528450012207, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8860169649124146, "num_tokens": 480264280.0, "step": 12590 }, { "epoch": 1.601704617733113, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.73328685760498, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8557548522949219, "num_tokens": 480300670.0, "step": 12591 }, { "epoch": 1.6018318280117034, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.742980003356934, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8681946992874146, "num_tokens": 480334460.0, "step": 12592 }, { "epoch": 1.601959038290294, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.68885326385498, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8787618279457092, "num_tokens": 480376066.0, "step": 12593 }, { "epoch": 1.6020862485688845, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.723489761352539, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8754463195800781, "num_tokens": 480417575.0, "step": 12594 }, { "epoch": 1.602213458847475, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.712945938110352, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8593489527702332, "num_tokens": 480454106.0, "step": 12595 }, { "epoch": 1.6023406691260655, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.72415542602539, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8575408458709717, "num_tokens": 480493031.0, "step": 12596 }, { "epoch": 1.6024678794046558, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.625371932983398, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8716614842414856, "num_tokens": 480528326.0, "step": 12597 }, { "epoch": 1.6025950896832464, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.698041915893555, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.877801775932312, "num_tokens": 480565399.0, "step": 12598 }, { "epoch": 1.602722299961837, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.756876945495605, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.8853136301040649, "num_tokens": 480600028.0, "step": 12599 }, { "epoch": 1.6028495102404274, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.659307479858398, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8681546449661255, "num_tokens": 480635759.0, "step": 12600 }, { "epoch": 1.602976720519018, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.708932876586914, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8748670816421509, "num_tokens": 480669528.0, "step": 12601 }, { "epoch": 1.6031039307976085, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.697818756103516, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.868304967880249, "num_tokens": 480709425.0, "step": 12602 }, { "epoch": 1.6032311410761988, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.717765808105469, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8558376431465149, "num_tokens": 480745010.0, "step": 12603 }, { "epoch": 1.6033583513547893, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.743199348449707, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.875984787940979, "num_tokens": 480780320.0, "step": 12604 }, { "epoch": 1.6034855616333799, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.694411277770996, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8765709400177002, "num_tokens": 480819335.0, "step": 12605 }, { "epoch": 1.6036127719119704, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.723593711853027, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.882308840751648, "num_tokens": 480852702.0, "step": 12606 }, { "epoch": 1.603739982190561, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.692174911499023, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8649985790252686, "num_tokens": 480890258.0, "step": 12607 }, { "epoch": 1.6038671924691514, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.705188751220703, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8697394132614136, "num_tokens": 480929555.0, "step": 12608 }, { "epoch": 1.603994402747742, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.719866752624512, "learning_rate": 1e-06, "loss": 0.3415, "mean_token_accuracy": 0.8924331665039062, "num_tokens": 480970295.0, "step": 12609 }, { "epoch": 1.6041216130263325, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.645936012268066, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8707695007324219, "num_tokens": 481011213.0, "step": 12610 }, { "epoch": 1.604248823304923, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.696805953979492, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8641359806060791, "num_tokens": 481045273.0, "step": 12611 }, { "epoch": 1.6043760335835135, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.690174102783203, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8503825664520264, "num_tokens": 481084567.0, "step": 12612 }, { "epoch": 1.604503243862104, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.723732948303223, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8713890314102173, "num_tokens": 481119455.0, "step": 12613 }, { "epoch": 1.6046304541406946, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.759066581726074, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8559830784797668, "num_tokens": 481154718.0, "step": 12614 }, { "epoch": 1.6047576644192851, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.627296447753906, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8641967177391052, "num_tokens": 481193716.0, "step": 12615 }, { "epoch": 1.6048848746978757, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.796265602111816, "learning_rate": 1e-06, "loss": 0.4869, "mean_token_accuracy": 0.8461601734161377, "num_tokens": 481231378.0, "step": 12616 }, { "epoch": 1.6050120849764662, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.729681968688965, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8684417009353638, "num_tokens": 481267907.0, "step": 12617 }, { "epoch": 1.6051392952550567, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.667793273925781, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8709990978240967, "num_tokens": 481309899.0, "step": 12618 }, { "epoch": 1.6052665055336472, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.697285652160645, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8791000247001648, "num_tokens": 481348845.0, "step": 12619 }, { "epoch": 1.6053937158122378, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.745946884155273, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8602694272994995, "num_tokens": 481384999.0, "step": 12620 }, { "epoch": 1.6055209260908283, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.726853370666504, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8712213039398193, "num_tokens": 481417287.0, "step": 12621 }, { "epoch": 1.6056481363694186, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.666868209838867, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8735767602920532, "num_tokens": 481455943.0, "step": 12622 }, { "epoch": 1.6057753466480091, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.777144432067871, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.88444983959198, "num_tokens": 481492017.0, "step": 12623 }, { "epoch": 1.6059025569265997, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.705327987670898, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8631929159164429, "num_tokens": 481535115.0, "step": 12624 }, { "epoch": 1.6060297672051902, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.724617004394531, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8546704053878784, "num_tokens": 481580048.0, "step": 12625 }, { "epoch": 1.6061569774837807, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.701057434082031, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.879850447177887, "num_tokens": 481616748.0, "step": 12626 }, { "epoch": 1.6062841877623713, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.750930786132812, "learning_rate": 1e-06, "loss": 0.4904, "mean_token_accuracy": 0.8411246538162231, "num_tokens": 481659123.0, "step": 12627 }, { "epoch": 1.6064113980409616, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.7233304977417, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8607326149940491, "num_tokens": 481704791.0, "step": 12628 }, { "epoch": 1.606538608319552, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.67735767364502, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8565698862075806, "num_tokens": 481742913.0, "step": 12629 }, { "epoch": 1.6066658185981426, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.644574165344238, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8561526536941528, "num_tokens": 481783549.0, "step": 12630 }, { "epoch": 1.6067930288767331, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.658702850341797, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.875047504901886, "num_tokens": 481818486.0, "step": 12631 }, { "epoch": 1.6069202391553237, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.684914588928223, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8696272969245911, "num_tokens": 481849969.0, "step": 12632 }, { "epoch": 1.6070474494339142, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.676645278930664, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8529049754142761, "num_tokens": 481886108.0, "step": 12633 }, { "epoch": 1.6071746597125047, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.705438613891602, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8808372616767883, "num_tokens": 481919389.0, "step": 12634 }, { "epoch": 1.6073018699910953, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.666769027709961, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8827651739120483, "num_tokens": 481958531.0, "step": 12635 }, { "epoch": 1.6074290802696858, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.746901512145996, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8655515909194946, "num_tokens": 481999864.0, "step": 12636 }, { "epoch": 1.6075562905482763, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.670013427734375, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.8921722173690796, "num_tokens": 482033890.0, "step": 12637 }, { "epoch": 1.6076835008268668, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.658535957336426, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8747934103012085, "num_tokens": 482071186.0, "step": 12638 }, { "epoch": 1.6078107111054574, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.738473892211914, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8678179979324341, "num_tokens": 482107218.0, "step": 12639 }, { "epoch": 1.607937921384048, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.670846939086914, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8654364943504333, "num_tokens": 482146101.0, "step": 12640 }, { "epoch": 1.6080651316626384, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.730611801147461, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8824158906936646, "num_tokens": 482178531.0, "step": 12641 }, { "epoch": 1.608192341941229, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.662026405334473, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8787005543708801, "num_tokens": 482217019.0, "step": 12642 }, { "epoch": 1.6083195522198195, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.742463111877441, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8705973625183105, "num_tokens": 482250547.0, "step": 12643 }, { "epoch": 1.60844676249841, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.730537414550781, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8650784492492676, "num_tokens": 482287009.0, "step": 12644 }, { "epoch": 1.6085739727770005, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.645670890808105, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8569048047065735, "num_tokens": 482327373.0, "step": 12645 }, { "epoch": 1.6087011830555908, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.783110618591309, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8580674529075623, "num_tokens": 482363409.0, "step": 12646 }, { "epoch": 1.6088283933341814, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.672690391540527, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8607367277145386, "num_tokens": 482396499.0, "step": 12647 }, { "epoch": 1.608955603612772, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.712594032287598, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8707084655761719, "num_tokens": 482435059.0, "step": 12648 }, { "epoch": 1.6090828138913624, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.708028793334961, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8627957701683044, "num_tokens": 482472793.0, "step": 12649 }, { "epoch": 1.609210024169953, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.740413665771484, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8714603781700134, "num_tokens": 482508857.0, "step": 12650 }, { "epoch": 1.6093372344485435, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.682965278625488, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8527981638908386, "num_tokens": 482548526.0, "step": 12651 }, { "epoch": 1.6094644447271338, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.682306289672852, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8624297976493835, "num_tokens": 482586461.0, "step": 12652 }, { "epoch": 1.6095916550057243, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.791664123535156, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8621070981025696, "num_tokens": 482624642.0, "step": 12653 }, { "epoch": 1.6097188652843148, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.678149223327637, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8684118986129761, "num_tokens": 482664096.0, "step": 12654 }, { "epoch": 1.6098460755629054, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.788453102111816, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8800696134567261, "num_tokens": 482701230.0, "step": 12655 }, { "epoch": 1.609973285841496, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.70828628540039, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.868873119354248, "num_tokens": 482741611.0, "step": 12656 }, { "epoch": 1.6101004961200864, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.770147323608398, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8791509866714478, "num_tokens": 482779482.0, "step": 12657 }, { "epoch": 1.610227706398677, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.744168281555176, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8770454525947571, "num_tokens": 482816279.0, "step": 12658 }, { "epoch": 1.6103549166772675, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.719481468200684, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8606774806976318, "num_tokens": 482859941.0, "step": 12659 }, { "epoch": 1.610482126955858, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.815940856933594, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8738780617713928, "num_tokens": 482894348.0, "step": 12660 }, { "epoch": 1.6106093372344485, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.687589645385742, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8632340431213379, "num_tokens": 482930882.0, "step": 12661 }, { "epoch": 1.610736547513039, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.72830867767334, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8518095016479492, "num_tokens": 482969515.0, "step": 12662 }, { "epoch": 1.6108637577916296, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.685919761657715, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8671762943267822, "num_tokens": 483008074.0, "step": 12663 }, { "epoch": 1.6109909680702201, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.679018020629883, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8619616627693176, "num_tokens": 483038920.0, "step": 12664 }, { "epoch": 1.6111181783488107, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.745026588439941, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8799863457679749, "num_tokens": 483075559.0, "step": 12665 }, { "epoch": 1.6112453886274012, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.676968574523926, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.872324526309967, "num_tokens": 483114378.0, "step": 12666 }, { "epoch": 1.6113725989059917, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.693739891052246, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8731131553649902, "num_tokens": 483147911.0, "step": 12667 }, { "epoch": 1.6114998091845822, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.712964057922363, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8838245868682861, "num_tokens": 483188275.0, "step": 12668 }, { "epoch": 1.6116270194631728, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.694158554077148, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.880137026309967, "num_tokens": 483230804.0, "step": 12669 }, { "epoch": 1.6117542297417633, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.758707046508789, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8726037740707397, "num_tokens": 483262368.0, "step": 12670 }, { "epoch": 1.6118814400203536, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.778035163879395, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.857994556427002, "num_tokens": 483300894.0, "step": 12671 }, { "epoch": 1.6120086502989441, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.82311725616455, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8758566379547119, "num_tokens": 483337377.0, "step": 12672 }, { "epoch": 1.6121358605775347, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.71680736541748, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8687217235565186, "num_tokens": 483376572.0, "step": 12673 }, { "epoch": 1.6122630708561252, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.75874137878418, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8573105335235596, "num_tokens": 483415584.0, "step": 12674 }, { "epoch": 1.6123902811347157, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.727785110473633, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.870327353477478, "num_tokens": 483459584.0, "step": 12675 }, { "epoch": 1.6125174914133062, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.711681365966797, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.863246500492096, "num_tokens": 483504260.0, "step": 12676 }, { "epoch": 1.6126447016918966, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.717430114746094, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.871815025806427, "num_tokens": 483547271.0, "step": 12677 }, { "epoch": 1.612771911970487, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.73010540008545, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8718066215515137, "num_tokens": 483582255.0, "step": 12678 }, { "epoch": 1.6128991222490776, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.775906562805176, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8653327226638794, "num_tokens": 483622290.0, "step": 12679 }, { "epoch": 1.6130263325276681, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.664941787719727, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8817017078399658, "num_tokens": 483657253.0, "step": 12680 }, { "epoch": 1.6131535428062587, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.79283618927002, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8736841678619385, "num_tokens": 483693062.0, "step": 12681 }, { "epoch": 1.6132807530848492, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.686264038085938, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8637924790382385, "num_tokens": 483735845.0, "step": 12682 }, { "epoch": 1.6134079633634397, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.709799766540527, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8727224469184875, "num_tokens": 483783833.0, "step": 12683 }, { "epoch": 1.6135351736420303, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.755498886108398, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8688039779663086, "num_tokens": 483820175.0, "step": 12684 }, { "epoch": 1.6136623839206208, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.731134414672852, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8809798359870911, "num_tokens": 483863454.0, "step": 12685 }, { "epoch": 1.6137895941992113, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.743924140930176, "learning_rate": 1e-06, "loss": 0.3436, "mean_token_accuracy": 0.8896815776824951, "num_tokens": 483900657.0, "step": 12686 }, { "epoch": 1.6139168044778018, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.775960922241211, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8630160093307495, "num_tokens": 483935912.0, "step": 12687 }, { "epoch": 1.6140440147563924, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.698753356933594, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8701447248458862, "num_tokens": 483968594.0, "step": 12688 }, { "epoch": 1.614171225034983, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.670778274536133, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8784005045890808, "num_tokens": 484010866.0, "step": 12689 }, { "epoch": 1.6142984353135734, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.652385711669922, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8629908561706543, "num_tokens": 484051355.0, "step": 12690 }, { "epoch": 1.614425645592164, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.798542022705078, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8803337812423706, "num_tokens": 484087049.0, "step": 12691 }, { "epoch": 1.6145528558707545, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.659689903259277, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8584691882133484, "num_tokens": 484124914.0, "step": 12692 }, { "epoch": 1.614680066149345, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.741410255432129, "learning_rate": 1e-06, "loss": 0.3557, "mean_token_accuracy": 0.8869196176528931, "num_tokens": 484154524.0, "step": 12693 }, { "epoch": 1.6148072764279355, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.681058883666992, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8676917552947998, "num_tokens": 484195091.0, "step": 12694 }, { "epoch": 1.6149344867065258, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.708251953125, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8721798658370972, "num_tokens": 484236787.0, "step": 12695 }, { "epoch": 1.6150616969851164, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.67910099029541, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8652479648590088, "num_tokens": 484281748.0, "step": 12696 }, { "epoch": 1.615188907263707, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.755545616149902, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8754004240036011, "num_tokens": 484321181.0, "step": 12697 }, { "epoch": 1.6153161175422974, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.68474006652832, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.867113471031189, "num_tokens": 484360476.0, "step": 12698 }, { "epoch": 1.615443327820888, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.729606628417969, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.870819091796875, "num_tokens": 484395065.0, "step": 12699 }, { "epoch": 1.6155705380994785, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.695991516113281, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8722549676895142, "num_tokens": 484439309.0, "step": 12700 }, { "epoch": 1.6156977483780688, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.651089668273926, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8705071806907654, "num_tokens": 484473018.0, "step": 12701 }, { "epoch": 1.6158249586566593, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.731237411499023, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8761686086654663, "num_tokens": 484518306.0, "step": 12702 }, { "epoch": 1.6159521689352498, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.741070747375488, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8565346002578735, "num_tokens": 484556348.0, "step": 12703 }, { "epoch": 1.6160793792138404, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.728326797485352, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8813948631286621, "num_tokens": 484593136.0, "step": 12704 }, { "epoch": 1.616206589492431, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.686759948730469, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8788742423057556, "num_tokens": 484628025.0, "step": 12705 }, { "epoch": 1.6163337997710214, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.708078384399414, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8753683567047119, "num_tokens": 484666703.0, "step": 12706 }, { "epoch": 1.616461010049612, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.691831588745117, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8796411156654358, "num_tokens": 484703417.0, "step": 12707 }, { "epoch": 1.6165882203282025, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.72021198272705, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8838541507720947, "num_tokens": 484741830.0, "step": 12708 }, { "epoch": 1.616715430606793, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.732475280761719, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8688608407974243, "num_tokens": 484780140.0, "step": 12709 }, { "epoch": 1.6168426408853835, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.743310928344727, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8683125376701355, "num_tokens": 484813856.0, "step": 12710 }, { "epoch": 1.616969851163974, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.668251037597656, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8793606162071228, "num_tokens": 484851383.0, "step": 12711 }, { "epoch": 1.6170970614425646, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.755614280700684, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8787826299667358, "num_tokens": 484884361.0, "step": 12712 }, { "epoch": 1.6172242717211551, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.635311126708984, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8500352501869202, "num_tokens": 484922106.0, "step": 12713 }, { "epoch": 1.6173514819997457, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.77591323852539, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.870613157749176, "num_tokens": 484963100.0, "step": 12714 }, { "epoch": 1.6174786922783362, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.70893669128418, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8812653422355652, "num_tokens": 485006896.0, "step": 12715 }, { "epoch": 1.6176059025569267, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.67874813079834, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8705333471298218, "num_tokens": 485045968.0, "step": 12716 }, { "epoch": 1.6177331128355172, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.692033767700195, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8675075173377991, "num_tokens": 485084736.0, "step": 12717 }, { "epoch": 1.6178603231141078, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.618549346923828, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.878699779510498, "num_tokens": 485123130.0, "step": 12718 }, { "epoch": 1.6179875333926983, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.698704719543457, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8667653203010559, "num_tokens": 485158072.0, "step": 12719 }, { "epoch": 1.6181147436712886, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.662155151367188, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8788883090019226, "num_tokens": 485196783.0, "step": 12720 }, { "epoch": 1.6182419539498791, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.773053169250488, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8762888312339783, "num_tokens": 485231433.0, "step": 12721 }, { "epoch": 1.6183691642284697, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.687921524047852, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8691354990005493, "num_tokens": 485269016.0, "step": 12722 }, { "epoch": 1.6184963745070602, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.75121784210205, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8595679998397827, "num_tokens": 485308293.0, "step": 12723 }, { "epoch": 1.6186235847856507, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.707314491271973, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8791185021400452, "num_tokens": 485340758.0, "step": 12724 }, { "epoch": 1.6187507950642412, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.745972633361816, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.8847417831420898, "num_tokens": 485384047.0, "step": 12725 }, { "epoch": 1.6188780053428315, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.750744819641113, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8740990161895752, "num_tokens": 485416909.0, "step": 12726 }, { "epoch": 1.619005215621422, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.720381736755371, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8708454966545105, "num_tokens": 485449532.0, "step": 12727 }, { "epoch": 1.6191324259000126, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.666962623596191, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8814612030982971, "num_tokens": 485491171.0, "step": 12728 }, { "epoch": 1.6192596361786031, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.715777397155762, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8654091358184814, "num_tokens": 485525511.0, "step": 12729 }, { "epoch": 1.6193868464571937, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.699843406677246, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8665664792060852, "num_tokens": 485562517.0, "step": 12730 }, { "epoch": 1.6195140567357842, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.721065521240234, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8721976280212402, "num_tokens": 485601058.0, "step": 12731 }, { "epoch": 1.6196412670143747, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.674291610717773, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8840636014938354, "num_tokens": 485634737.0, "step": 12732 }, { "epoch": 1.6197684772929652, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.704931259155273, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8743196725845337, "num_tokens": 485671300.0, "step": 12733 }, { "epoch": 1.6198956875715558, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.649547576904297, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8782299757003784, "num_tokens": 485708712.0, "step": 12734 }, { "epoch": 1.6200228978501463, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.650938034057617, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8833125233650208, "num_tokens": 485746820.0, "step": 12735 }, { "epoch": 1.6201501081287368, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.63976764678955, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8712077736854553, "num_tokens": 485785921.0, "step": 12736 }, { "epoch": 1.6202773184073274, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.722800254821777, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.871290922164917, "num_tokens": 485825766.0, "step": 12737 }, { "epoch": 1.6204045286859179, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.690511703491211, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8643591403961182, "num_tokens": 485862288.0, "step": 12738 }, { "epoch": 1.6205317389645084, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.725127220153809, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8644691705703735, "num_tokens": 485903582.0, "step": 12739 }, { "epoch": 1.620658949243099, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.671234130859375, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8673468828201294, "num_tokens": 485950931.0, "step": 12740 }, { "epoch": 1.6207861595216895, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.724273681640625, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8727917671203613, "num_tokens": 485984185.0, "step": 12741 }, { "epoch": 1.62091336980028, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.671685218811035, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8763983249664307, "num_tokens": 486018796.0, "step": 12742 }, { "epoch": 1.6210405800788705, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.776394844055176, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8576619029045105, "num_tokens": 486055328.0, "step": 12743 }, { "epoch": 1.6211677903574608, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.660394668579102, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8652431964874268, "num_tokens": 486092340.0, "step": 12744 }, { "epoch": 1.6212950006360514, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.76102066040039, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8578299283981323, "num_tokens": 486123195.0, "step": 12745 }, { "epoch": 1.621422210914642, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.677922248840332, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8803524971008301, "num_tokens": 486167145.0, "step": 12746 }, { "epoch": 1.6215494211932324, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.673632621765137, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8623328804969788, "num_tokens": 486207925.0, "step": 12747 }, { "epoch": 1.621676631471823, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.726157188415527, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8753610849380493, "num_tokens": 486245446.0, "step": 12748 }, { "epoch": 1.6218038417504135, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.752370834350586, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.871198832988739, "num_tokens": 486280461.0, "step": 12749 }, { "epoch": 1.6219310520290038, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.78054428100586, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8591499924659729, "num_tokens": 486323926.0, "step": 12750 }, { "epoch": 1.6220582623075943, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.704475402832031, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8638149499893188, "num_tokens": 486360016.0, "step": 12751 }, { "epoch": 1.6221854725861848, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.784130096435547, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8738048076629639, "num_tokens": 486394994.0, "step": 12752 }, { "epoch": 1.6223126828647754, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.740294456481934, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8694964647293091, "num_tokens": 486436909.0, "step": 12753 }, { "epoch": 1.622439893143366, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.779664039611816, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8768379092216492, "num_tokens": 486473201.0, "step": 12754 }, { "epoch": 1.6225671034219564, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.687399864196777, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8519039154052734, "num_tokens": 486510733.0, "step": 12755 }, { "epoch": 1.622694313700547, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.762916564941406, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8862681984901428, "num_tokens": 486546666.0, "step": 12756 }, { "epoch": 1.6228215239791375, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.80384349822998, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8519572019577026, "num_tokens": 486585639.0, "step": 12757 }, { "epoch": 1.622948734257728, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.753975868225098, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.87496417760849, "num_tokens": 486623028.0, "step": 12758 }, { "epoch": 1.6230759445363185, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.787841796875, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8657858371734619, "num_tokens": 486655906.0, "step": 12759 }, { "epoch": 1.623203154814909, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.716470718383789, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.867472767829895, "num_tokens": 486689461.0, "step": 12760 }, { "epoch": 1.6233303650934996, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.74162769317627, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8545933365821838, "num_tokens": 486728373.0, "step": 12761 }, { "epoch": 1.6234575753720901, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.723153114318848, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8804162740707397, "num_tokens": 486769955.0, "step": 12762 }, { "epoch": 1.6235847856506807, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.768227577209473, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.87404465675354, "num_tokens": 486806468.0, "step": 12763 }, { "epoch": 1.6237119959292712, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.777112007141113, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8574650287628174, "num_tokens": 486846526.0, "step": 12764 }, { "epoch": 1.6238392062078617, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.725361824035645, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8879811763763428, "num_tokens": 486886323.0, "step": 12765 }, { "epoch": 1.6239664164864522, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.830633163452148, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8521246314048767, "num_tokens": 486918099.0, "step": 12766 }, { "epoch": 1.6240936267650428, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.685147285461426, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8850443959236145, "num_tokens": 486953789.0, "step": 12767 }, { "epoch": 1.6242208370436333, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.715985298156738, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8701429963111877, "num_tokens": 486996165.0, "step": 12768 }, { "epoch": 1.6243480473222236, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.721652030944824, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8816541433334351, "num_tokens": 487028249.0, "step": 12769 }, { "epoch": 1.6244752576008141, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.75206470489502, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8714679479598999, "num_tokens": 487070099.0, "step": 12770 }, { "epoch": 1.6246024678794047, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.725469589233398, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8740421533584595, "num_tokens": 487107117.0, "step": 12771 }, { "epoch": 1.6247296781579952, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.709694862365723, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.858778715133667, "num_tokens": 487146024.0, "step": 12772 }, { "epoch": 1.6248568884365857, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.70418643951416, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8581218719482422, "num_tokens": 487189864.0, "step": 12773 }, { "epoch": 1.6249840987151762, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.728276252746582, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8735089898109436, "num_tokens": 487230307.0, "step": 12774 }, { "epoch": 1.6251113089937665, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.827799797058105, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8738757371902466, "num_tokens": 487261387.0, "step": 12775 }, { "epoch": 1.625238519272357, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.714641571044922, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8598734140396118, "num_tokens": 487298611.0, "step": 12776 }, { "epoch": 1.6253657295509476, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.773284912109375, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8702902793884277, "num_tokens": 487338172.0, "step": 12777 }, { "epoch": 1.6254929398295381, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.717486381530762, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8717955350875854, "num_tokens": 487375126.0, "step": 12778 }, { "epoch": 1.6256201501081287, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.715841293334961, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8815884590148926, "num_tokens": 487416539.0, "step": 12779 }, { "epoch": 1.6257473603867192, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.728211402893066, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8734963536262512, "num_tokens": 487448429.0, "step": 12780 }, { "epoch": 1.6258745706653097, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.739126205444336, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8706094026565552, "num_tokens": 487486054.0, "step": 12781 }, { "epoch": 1.6260017809439002, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.643511772155762, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8761926293373108, "num_tokens": 487523771.0, "step": 12782 }, { "epoch": 1.6261289912224908, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.74641227722168, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8754122257232666, "num_tokens": 487561334.0, "step": 12783 }, { "epoch": 1.6262562015010813, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.708487510681152, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8718605041503906, "num_tokens": 487606382.0, "step": 12784 }, { "epoch": 1.6263834117796718, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.729011535644531, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8662227988243103, "num_tokens": 487647640.0, "step": 12785 }, { "epoch": 1.6265106220582624, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.827991485595703, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8669472336769104, "num_tokens": 487682604.0, "step": 12786 }, { "epoch": 1.6266378323368529, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.758384704589844, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8741375207901001, "num_tokens": 487718271.0, "step": 12787 }, { "epoch": 1.6267650426154434, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.760997772216797, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8578723669052124, "num_tokens": 487752872.0, "step": 12788 }, { "epoch": 1.626892252894034, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.715435028076172, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.861059844493866, "num_tokens": 487798309.0, "step": 12789 }, { "epoch": 1.6270194631726245, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.808711051940918, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8595003485679626, "num_tokens": 487834052.0, "step": 12790 }, { "epoch": 1.627146673451215, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.801541328430176, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8588430881500244, "num_tokens": 487868662.0, "step": 12791 }, { "epoch": 1.6272738837298055, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.786645889282227, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8678814172744751, "num_tokens": 487907425.0, "step": 12792 }, { "epoch": 1.6274010940083958, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.823341369628906, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8788734674453735, "num_tokens": 487942979.0, "step": 12793 }, { "epoch": 1.6275283042869864, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.818364143371582, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8751622438430786, "num_tokens": 487982402.0, "step": 12794 }, { "epoch": 1.6276555145655769, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.844334602355957, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8776403665542603, "num_tokens": 488020071.0, "step": 12795 }, { "epoch": 1.6277827248441674, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.73934268951416, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8573276996612549, "num_tokens": 488057457.0, "step": 12796 }, { "epoch": 1.627909935122758, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.78186321258545, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8670650720596313, "num_tokens": 488094002.0, "step": 12797 }, { "epoch": 1.6280371454013485, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.756359100341797, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8561508655548096, "num_tokens": 488134995.0, "step": 12798 }, { "epoch": 1.6281643556799388, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.810495376586914, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.873651385307312, "num_tokens": 488169942.0, "step": 12799 }, { "epoch": 1.6282915659585293, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.682394027709961, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8727409839630127, "num_tokens": 488210852.0, "step": 12800 }, { "epoch": 1.6284187762371198, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.7778959274292, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8739036321640015, "num_tokens": 488253067.0, "step": 12801 }, { "epoch": 1.6285459865157104, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.798439979553223, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8782566785812378, "num_tokens": 488293498.0, "step": 12802 }, { "epoch": 1.628673196794301, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.775333404541016, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8717600703239441, "num_tokens": 488333931.0, "step": 12803 }, { "epoch": 1.6288004070728914, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.798928260803223, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8745877742767334, "num_tokens": 488371724.0, "step": 12804 }, { "epoch": 1.628927617351482, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.813651084899902, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8650447130203247, "num_tokens": 488415171.0, "step": 12805 }, { "epoch": 1.6290548276300725, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.757672309875488, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8646191358566284, "num_tokens": 488451672.0, "step": 12806 }, { "epoch": 1.629182037908663, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.763914108276367, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8843293786048889, "num_tokens": 488492657.0, "step": 12807 }, { "epoch": 1.6293092481872535, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.724254608154297, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8773505091667175, "num_tokens": 488530681.0, "step": 12808 }, { "epoch": 1.629436458465844, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.779603004455566, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8804866671562195, "num_tokens": 488560542.0, "step": 12809 }, { "epoch": 1.6295636687444346, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.736786842346191, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8721592426300049, "num_tokens": 488597491.0, "step": 12810 }, { "epoch": 1.6296908790230251, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.81708812713623, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8784961700439453, "num_tokens": 488634556.0, "step": 12811 }, { "epoch": 1.6298180893016156, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.826348304748535, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8677798509597778, "num_tokens": 488668174.0, "step": 12812 }, { "epoch": 1.6299452995802062, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.766119003295898, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8729824423789978, "num_tokens": 488705994.0, "step": 12813 }, { "epoch": 1.6300725098587967, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.83135986328125, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8603805899620056, "num_tokens": 488746908.0, "step": 12814 }, { "epoch": 1.6301997201373872, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.752283096313477, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8710063695907593, "num_tokens": 488785607.0, "step": 12815 }, { "epoch": 1.6303269304159778, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.733490943908691, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8691163063049316, "num_tokens": 488823831.0, "step": 12816 }, { "epoch": 1.630454140694568, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.77358341217041, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8784106969833374, "num_tokens": 488859944.0, "step": 12817 }, { "epoch": 1.6305813509731586, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.773821830749512, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8559601902961731, "num_tokens": 488903764.0, "step": 12818 }, { "epoch": 1.6307085612517491, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.777318000793457, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.860526442527771, "num_tokens": 488946890.0, "step": 12819 }, { "epoch": 1.6308357715303397, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.8375244140625, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8676880598068237, "num_tokens": 488981901.0, "step": 12820 }, { "epoch": 1.6309629818089302, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.792475700378418, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8731327056884766, "num_tokens": 489024180.0, "step": 12821 }, { "epoch": 1.6310901920875207, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.78520679473877, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8736006021499634, "num_tokens": 489070971.0, "step": 12822 }, { "epoch": 1.6312174023661112, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.748312950134277, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8803869485855103, "num_tokens": 489111636.0, "step": 12823 }, { "epoch": 1.6313446126447015, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.76044750213623, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8687318563461304, "num_tokens": 489148677.0, "step": 12824 }, { "epoch": 1.631471822923292, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.718154907226562, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8583943843841553, "num_tokens": 489186222.0, "step": 12825 }, { "epoch": 1.6315990332018826, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.872696876525879, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8450108766555786, "num_tokens": 489223173.0, "step": 12826 }, { "epoch": 1.6317262434804731, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.720297813415527, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8706295490264893, "num_tokens": 489259236.0, "step": 12827 }, { "epoch": 1.6318534537590637, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.811105728149414, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8635470867156982, "num_tokens": 489298449.0, "step": 12828 }, { "epoch": 1.6319806640376542, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.749163627624512, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8500375747680664, "num_tokens": 489334800.0, "step": 12829 }, { "epoch": 1.6321078743162447, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.734237670898438, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8762762546539307, "num_tokens": 489378510.0, "step": 12830 }, { "epoch": 1.6322350845948352, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.76921272277832, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8592206835746765, "num_tokens": 489419267.0, "step": 12831 }, { "epoch": 1.6323622948734258, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.81678581237793, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.873360276222229, "num_tokens": 489454772.0, "step": 12832 }, { "epoch": 1.6324895051520163, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.748802185058594, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8649296760559082, "num_tokens": 489490146.0, "step": 12833 }, { "epoch": 1.6326167154306068, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.810161590576172, "learning_rate": 1e-06, "loss": 0.5001, "mean_token_accuracy": 0.8410667181015015, "num_tokens": 489532754.0, "step": 12834 }, { "epoch": 1.6327439257091974, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.8215970993042, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8680251836776733, "num_tokens": 489570378.0, "step": 12835 }, { "epoch": 1.6328711359877879, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.752004623413086, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8624051809310913, "num_tokens": 489604562.0, "step": 12836 }, { "epoch": 1.6329983462663784, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.786046981811523, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8778377771377563, "num_tokens": 489644060.0, "step": 12837 }, { "epoch": 1.633125556544969, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.76998233795166, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8717173337936401, "num_tokens": 489682693.0, "step": 12838 }, { "epoch": 1.6332527668235595, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.767793655395508, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8785544633865356, "num_tokens": 489720786.0, "step": 12839 }, { "epoch": 1.63337997710215, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.783394813537598, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.876441240310669, "num_tokens": 489758773.0, "step": 12840 }, { "epoch": 1.6335071873807405, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.757749557495117, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.870607852935791, "num_tokens": 489793436.0, "step": 12841 }, { "epoch": 1.6336343976593308, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.789217948913574, "learning_rate": 1e-06, "loss": 0.3439, "mean_token_accuracy": 0.8889858722686768, "num_tokens": 489828634.0, "step": 12842 }, { "epoch": 1.6337616079379214, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.711370468139648, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.873956561088562, "num_tokens": 489866335.0, "step": 12843 }, { "epoch": 1.6338888182165119, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.805249214172363, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8780727386474609, "num_tokens": 489903059.0, "step": 12844 }, { "epoch": 1.6340160284951024, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.711174011230469, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8774610757827759, "num_tokens": 489940186.0, "step": 12845 }, { "epoch": 1.634143238773693, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.654308319091797, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8821868896484375, "num_tokens": 489984732.0, "step": 12846 }, { "epoch": 1.6342704490522835, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.82826042175293, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8717601299285889, "num_tokens": 490016699.0, "step": 12847 }, { "epoch": 1.6343976593308738, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.713517189025879, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8625861406326294, "num_tokens": 490051351.0, "step": 12848 }, { "epoch": 1.6345248696094643, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.746176719665527, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8791248798370361, "num_tokens": 490089668.0, "step": 12849 }, { "epoch": 1.6346520798880548, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.696606636047363, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.871245265007019, "num_tokens": 490129651.0, "step": 12850 }, { "epoch": 1.6347792901666454, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.765242576599121, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8630480766296387, "num_tokens": 490168441.0, "step": 12851 }, { "epoch": 1.6349065004452359, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.78300952911377, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.880827784538269, "num_tokens": 490204754.0, "step": 12852 }, { "epoch": 1.6350337107238264, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.717849731445312, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8853607773780823, "num_tokens": 490244178.0, "step": 12853 }, { "epoch": 1.635160921002417, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.794672012329102, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8667454719543457, "num_tokens": 490281161.0, "step": 12854 }, { "epoch": 1.6352881312810075, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.764701843261719, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8716857433319092, "num_tokens": 490314892.0, "step": 12855 }, { "epoch": 1.635415341559598, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.743334770202637, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8597885966300964, "num_tokens": 490352600.0, "step": 12856 }, { "epoch": 1.6355425518381885, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.808491706848145, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8819149732589722, "num_tokens": 490384487.0, "step": 12857 }, { "epoch": 1.635669762116779, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.747424125671387, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8509852886199951, "num_tokens": 490422199.0, "step": 12858 }, { "epoch": 1.6357969723953696, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.72463321685791, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8633765578269958, "num_tokens": 490458058.0, "step": 12859 }, { "epoch": 1.6359241826739601, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.718915939331055, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8814241886138916, "num_tokens": 490498447.0, "step": 12860 }, { "epoch": 1.6360513929525506, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.7178316116333, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8578169345855713, "num_tokens": 490538502.0, "step": 12861 }, { "epoch": 1.6361786032311412, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.71668529510498, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8748384118080139, "num_tokens": 490575752.0, "step": 12862 }, { "epoch": 1.6363058135097317, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.767433166503906, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8703165054321289, "num_tokens": 490613502.0, "step": 12863 }, { "epoch": 1.6364330237883222, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.709078788757324, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8646076321601868, "num_tokens": 490653820.0, "step": 12864 }, { "epoch": 1.6365602340669128, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.793241500854492, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8552057147026062, "num_tokens": 490693332.0, "step": 12865 }, { "epoch": 1.636687444345503, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.751893997192383, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8691641092300415, "num_tokens": 490733135.0, "step": 12866 }, { "epoch": 1.6368146546240936, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.744084358215332, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.873749852180481, "num_tokens": 490775505.0, "step": 12867 }, { "epoch": 1.6369418649026841, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.730966567993164, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8702178001403809, "num_tokens": 490817783.0, "step": 12868 }, { "epoch": 1.6370690751812746, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.754743576049805, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8646743297576904, "num_tokens": 490849464.0, "step": 12869 }, { "epoch": 1.6371962854598652, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.710883140563965, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8643850088119507, "num_tokens": 490891925.0, "step": 12870 }, { "epoch": 1.6373234957384557, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.82895565032959, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8641365766525269, "num_tokens": 490926221.0, "step": 12871 }, { "epoch": 1.6374507060170462, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.75137996673584, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8612366914749146, "num_tokens": 490964141.0, "step": 12872 }, { "epoch": 1.6375779162956365, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.741989135742188, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.865974485874176, "num_tokens": 490997917.0, "step": 12873 }, { "epoch": 1.637705126574227, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.793802261352539, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8836768865585327, "num_tokens": 491029244.0, "step": 12874 }, { "epoch": 1.6378323368528176, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.741435050964355, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8698455095291138, "num_tokens": 491066591.0, "step": 12875 }, { "epoch": 1.6379595471314081, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.712579727172852, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8687006831169128, "num_tokens": 491106546.0, "step": 12876 }, { "epoch": 1.6380867574099987, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.737388610839844, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8632715940475464, "num_tokens": 491145512.0, "step": 12877 }, { "epoch": 1.6382139676885892, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.769558906555176, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8684235215187073, "num_tokens": 491185624.0, "step": 12878 }, { "epoch": 1.6383411779671797, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.689813613891602, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8638738989830017, "num_tokens": 491220982.0, "step": 12879 }, { "epoch": 1.6384683882457702, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.87255573272705, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8721112608909607, "num_tokens": 491262501.0, "step": 12880 }, { "epoch": 1.6385955985243608, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.764623641967773, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8814988732337952, "num_tokens": 491298078.0, "step": 12881 }, { "epoch": 1.6387228088029513, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.708796501159668, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8769131898880005, "num_tokens": 491336203.0, "step": 12882 }, { "epoch": 1.6388500190815418, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.805608749389648, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8589156270027161, "num_tokens": 491374723.0, "step": 12883 }, { "epoch": 1.6389772293601323, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.761112213134766, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8710840940475464, "num_tokens": 491410546.0, "step": 12884 }, { "epoch": 1.6391044396387229, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.759161949157715, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8660749197006226, "num_tokens": 491450642.0, "step": 12885 }, { "epoch": 1.6392316499173134, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.772022247314453, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.867005467414856, "num_tokens": 491489893.0, "step": 12886 }, { "epoch": 1.639358860195904, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.635074615478516, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8614569306373596, "num_tokens": 491535870.0, "step": 12887 }, { "epoch": 1.6394860704744945, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.751670837402344, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8685673475265503, "num_tokens": 491575988.0, "step": 12888 }, { "epoch": 1.639613280753085, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.722296714782715, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8722074031829834, "num_tokens": 491616822.0, "step": 12889 }, { "epoch": 1.6397404910316755, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.6715087890625, "learning_rate": 1e-06, "loss": 0.3458, "mean_token_accuracy": 0.8895602226257324, "num_tokens": 491654132.0, "step": 12890 }, { "epoch": 1.6398677013102658, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.759897232055664, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.857062041759491, "num_tokens": 491697215.0, "step": 12891 }, { "epoch": 1.6399949115888564, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.74129867553711, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8488597869873047, "num_tokens": 491731338.0, "step": 12892 }, { "epoch": 1.6401221218674469, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.759942054748535, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.867923378944397, "num_tokens": 491770808.0, "step": 12893 }, { "epoch": 1.6402493321460374, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.744524955749512, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8772416710853577, "num_tokens": 491804167.0, "step": 12894 }, { "epoch": 1.640376542424628, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.80798625946045, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8708636164665222, "num_tokens": 491840094.0, "step": 12895 }, { "epoch": 1.6405037527032185, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.683714866638184, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8556404113769531, "num_tokens": 491881643.0, "step": 12896 }, { "epoch": 1.6406309629818088, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.697498321533203, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.856484055519104, "num_tokens": 491920951.0, "step": 12897 }, { "epoch": 1.6407581732603993, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.721120834350586, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8659531474113464, "num_tokens": 491954599.0, "step": 12898 }, { "epoch": 1.6408853835389898, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.683273315429688, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8704043030738831, "num_tokens": 491995375.0, "step": 12899 }, { "epoch": 1.6410125938175804, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.667190551757812, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8647951483726501, "num_tokens": 492035297.0, "step": 12900 }, { "epoch": 1.6411398040961709, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.80218505859375, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.870013415813446, "num_tokens": 492071003.0, "step": 12901 }, { "epoch": 1.6412670143747614, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.783889770507812, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8692406415939331, "num_tokens": 492106046.0, "step": 12902 }, { "epoch": 1.641394224653352, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.73959732055664, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8815164566040039, "num_tokens": 492147252.0, "step": 12903 }, { "epoch": 1.6415214349319425, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.802857398986816, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8614023923873901, "num_tokens": 492188185.0, "step": 12904 }, { "epoch": 1.641648645210533, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.770995140075684, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8750447034835815, "num_tokens": 492219630.0, "step": 12905 }, { "epoch": 1.6417758554891235, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.804669380187988, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8718427419662476, "num_tokens": 492251955.0, "step": 12906 }, { "epoch": 1.641903065767714, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.725242614746094, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8719387650489807, "num_tokens": 492293484.0, "step": 12907 }, { "epoch": 1.6420302760463046, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.802192687988281, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8544167280197144, "num_tokens": 492331169.0, "step": 12908 }, { "epoch": 1.642157486324895, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.75314998626709, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8769446611404419, "num_tokens": 492366071.0, "step": 12909 }, { "epoch": 1.6422846966034856, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.797809600830078, "learning_rate": 1e-06, "loss": 0.3421, "mean_token_accuracy": 0.889613687992096, "num_tokens": 492396820.0, "step": 12910 }, { "epoch": 1.6424119068820762, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.796893119812012, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8761181831359863, "num_tokens": 492435282.0, "step": 12911 }, { "epoch": 1.6425391171606667, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.78914737701416, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.877254843711853, "num_tokens": 492470195.0, "step": 12912 }, { "epoch": 1.6426663274392572, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.749568939208984, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8634164929389954, "num_tokens": 492514265.0, "step": 12913 }, { "epoch": 1.6427935377178478, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.729196548461914, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8720324039459229, "num_tokens": 492550348.0, "step": 12914 }, { "epoch": 1.642920747996438, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.71069622039795, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8727993369102478, "num_tokens": 492590648.0, "step": 12915 }, { "epoch": 1.6430479582750286, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.740011215209961, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8712999820709229, "num_tokens": 492629523.0, "step": 12916 }, { "epoch": 1.6431751685536191, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.704142570495605, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8654302358627319, "num_tokens": 492668392.0, "step": 12917 }, { "epoch": 1.6433023788322096, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.806353569030762, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8596000671386719, "num_tokens": 492696073.0, "step": 12918 }, { "epoch": 1.6434295891108002, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.74435043334961, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8725109696388245, "num_tokens": 492728590.0, "step": 12919 }, { "epoch": 1.6435567993893907, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.723569869995117, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8623912334442139, "num_tokens": 492767290.0, "step": 12920 }, { "epoch": 1.6436840096679812, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.72424602508545, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.881083071231842, "num_tokens": 492801019.0, "step": 12921 }, { "epoch": 1.6438112199465715, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.768948554992676, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8688580393791199, "num_tokens": 492841220.0, "step": 12922 }, { "epoch": 1.643938430225162, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.718270301818848, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8592907190322876, "num_tokens": 492885076.0, "step": 12923 }, { "epoch": 1.6440656405037526, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.75311279296875, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8658822178840637, "num_tokens": 492918554.0, "step": 12924 }, { "epoch": 1.6441928507823431, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.683828353881836, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8858734965324402, "num_tokens": 492961890.0, "step": 12925 }, { "epoch": 1.6443200610609336, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.69743537902832, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8800503015518188, "num_tokens": 492997768.0, "step": 12926 }, { "epoch": 1.6444472713395242, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.768628120422363, "learning_rate": 1e-06, "loss": 0.3338, "mean_token_accuracy": 0.8942449688911438, "num_tokens": 493037017.0, "step": 12927 }, { "epoch": 1.6445744816181147, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.770309448242188, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8794738054275513, "num_tokens": 493074579.0, "step": 12928 }, { "epoch": 1.6447016918967052, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.756113052368164, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8710261583328247, "num_tokens": 493111249.0, "step": 12929 }, { "epoch": 1.6448289021752958, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.742476463317871, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8546123504638672, "num_tokens": 493150723.0, "step": 12930 }, { "epoch": 1.6449561124538863, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.754838943481445, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8682832717895508, "num_tokens": 493189530.0, "step": 12931 }, { "epoch": 1.6450833227324768, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.717820167541504, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.877147376537323, "num_tokens": 493231369.0, "step": 12932 }, { "epoch": 1.6452105330110673, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.764517784118652, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8565988540649414, "num_tokens": 493278430.0, "step": 12933 }, { "epoch": 1.6453377432896579, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.804430961608887, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8751586079597473, "num_tokens": 493312258.0, "step": 12934 }, { "epoch": 1.6454649535682484, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.780381202697754, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8684788942337036, "num_tokens": 493347725.0, "step": 12935 }, { "epoch": 1.645592163846839, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.76379680633545, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8745377063751221, "num_tokens": 493388119.0, "step": 12936 }, { "epoch": 1.6457193741254295, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.75438404083252, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8718134164810181, "num_tokens": 493424607.0, "step": 12937 }, { "epoch": 1.64584658440402, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.771544456481934, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.84703528881073, "num_tokens": 493459546.0, "step": 12938 }, { "epoch": 1.6459737946826105, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.715551376342773, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8737499713897705, "num_tokens": 493494162.0, "step": 12939 }, { "epoch": 1.6461010049612008, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.803362846374512, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8674831390380859, "num_tokens": 493537523.0, "step": 12940 }, { "epoch": 1.6462282152397913, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.7045316696167, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8692058324813843, "num_tokens": 493580065.0, "step": 12941 }, { "epoch": 1.6463554255183819, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.724045753479004, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8773422241210938, "num_tokens": 493626717.0, "step": 12942 }, { "epoch": 1.6464826357969724, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.709297180175781, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8761529922485352, "num_tokens": 493666266.0, "step": 12943 }, { "epoch": 1.646609846075563, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.821471214294434, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8604906797409058, "num_tokens": 493703184.0, "step": 12944 }, { "epoch": 1.6467370563541535, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.661009788513184, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8852518796920776, "num_tokens": 493744061.0, "step": 12945 }, { "epoch": 1.6468642666327438, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.741493225097656, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8701124787330627, "num_tokens": 493783585.0, "step": 12946 }, { "epoch": 1.6469914769113343, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.819488525390625, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.876820981502533, "num_tokens": 493827781.0, "step": 12947 }, { "epoch": 1.6471186871899248, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.67171573638916, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8754269480705261, "num_tokens": 493864418.0, "step": 12948 }, { "epoch": 1.6472458974685154, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.780855178833008, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8691566586494446, "num_tokens": 493895226.0, "step": 12949 }, { "epoch": 1.6473731077471059, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.696486473083496, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8614991903305054, "num_tokens": 493935210.0, "step": 12950 }, { "epoch": 1.6475003180256964, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.739753723144531, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8694473505020142, "num_tokens": 493966368.0, "step": 12951 }, { "epoch": 1.647627528304287, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.664580345153809, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8637258410453796, "num_tokens": 494004304.0, "step": 12952 }, { "epoch": 1.6477547385828775, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.740839958190918, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8767646551132202, "num_tokens": 494044758.0, "step": 12953 }, { "epoch": 1.647881948861468, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.713028907775879, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8544345498085022, "num_tokens": 494080624.0, "step": 12954 }, { "epoch": 1.6480091591400585, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.697171211242676, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8759599328041077, "num_tokens": 494116215.0, "step": 12955 }, { "epoch": 1.648136369418649, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.768746376037598, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8694472908973694, "num_tokens": 494150317.0, "step": 12956 }, { "epoch": 1.6482635796972396, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.748205184936523, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8844997882843018, "num_tokens": 494187273.0, "step": 12957 }, { "epoch": 1.64839078997583, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.795509338378906, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8665347099304199, "num_tokens": 494226921.0, "step": 12958 }, { "epoch": 1.6485180002544206, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.7283353805542, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8765732645988464, "num_tokens": 494263141.0, "step": 12959 }, { "epoch": 1.6486452105330112, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.803722381591797, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8667222261428833, "num_tokens": 494307925.0, "step": 12960 }, { "epoch": 1.6487724208116017, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.74217700958252, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8639750480651855, "num_tokens": 494353428.0, "step": 12961 }, { "epoch": 1.6488996310901922, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.781173706054688, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8582303524017334, "num_tokens": 494383559.0, "step": 12962 }, { "epoch": 1.6490268413687827, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.803287506103516, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8730505704879761, "num_tokens": 494419564.0, "step": 12963 }, { "epoch": 1.649154051647373, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.721346855163574, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.873126745223999, "num_tokens": 494466886.0, "step": 12964 }, { "epoch": 1.6492812619259636, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.788972854614258, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8531877994537354, "num_tokens": 494506672.0, "step": 12965 }, { "epoch": 1.649408472204554, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.838180541992188, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.869539201259613, "num_tokens": 494541881.0, "step": 12966 }, { "epoch": 1.6495356824831446, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.757001876831055, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8749587535858154, "num_tokens": 494583546.0, "step": 12967 }, { "epoch": 1.6496628927617352, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.754916191101074, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8589766025543213, "num_tokens": 494616625.0, "step": 12968 }, { "epoch": 1.6497901030403257, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.709975242614746, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8832892179489136, "num_tokens": 494654590.0, "step": 12969 }, { "epoch": 1.6499173133189162, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.704108238220215, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8837733864784241, "num_tokens": 494692471.0, "step": 12970 }, { "epoch": 1.6500445235975065, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.789817810058594, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8622029423713684, "num_tokens": 494730347.0, "step": 12971 }, { "epoch": 1.650171733876097, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.744670867919922, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8585637807846069, "num_tokens": 494763753.0, "step": 12972 }, { "epoch": 1.6502989441546876, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.737954139709473, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8650892376899719, "num_tokens": 494800416.0, "step": 12973 }, { "epoch": 1.6504261544332781, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.725549697875977, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8686822652816772, "num_tokens": 494842733.0, "step": 12974 }, { "epoch": 1.6505533647118686, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.770014762878418, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8634658455848694, "num_tokens": 494878661.0, "step": 12975 }, { "epoch": 1.6506805749904592, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.793830871582031, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8625697493553162, "num_tokens": 494918142.0, "step": 12976 }, { "epoch": 1.6508077852690497, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.846715927124023, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8694202303886414, "num_tokens": 494956947.0, "step": 12977 }, { "epoch": 1.6509349955476402, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.82528305053711, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8654221892356873, "num_tokens": 494989850.0, "step": 12978 }, { "epoch": 1.6510622058262308, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.778807640075684, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8622438907623291, "num_tokens": 495026770.0, "step": 12979 }, { "epoch": 1.6511894161048213, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.870779991149902, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8814163208007812, "num_tokens": 495065009.0, "step": 12980 }, { "epoch": 1.6513166263834118, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.775014877319336, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8707939386367798, "num_tokens": 495096914.0, "step": 12981 }, { "epoch": 1.6514438366620023, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.86214828491211, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8740376234054565, "num_tokens": 495138348.0, "step": 12982 }, { "epoch": 1.6515710469405929, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.742508888244629, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8633428812026978, "num_tokens": 495177026.0, "step": 12983 }, { "epoch": 1.6516982572191834, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.751581192016602, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8714842796325684, "num_tokens": 495213850.0, "step": 12984 }, { "epoch": 1.651825467497774, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.811376571655273, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8781898021697998, "num_tokens": 495257213.0, "step": 12985 }, { "epoch": 1.6519526777763645, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.796930313110352, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.860355019569397, "num_tokens": 495300018.0, "step": 12986 }, { "epoch": 1.652079888054955, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.743523597717285, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8678804039955139, "num_tokens": 495336471.0, "step": 12987 }, { "epoch": 1.6522070983335455, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.786042213439941, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8735674023628235, "num_tokens": 495374680.0, "step": 12988 }, { "epoch": 1.6523343086121358, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.755289077758789, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8727591037750244, "num_tokens": 495414096.0, "step": 12989 }, { "epoch": 1.6524615188907263, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.791776657104492, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8571237921714783, "num_tokens": 495452955.0, "step": 12990 }, { "epoch": 1.6525887291693169, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.82800579071045, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8669877052307129, "num_tokens": 495490091.0, "step": 12991 }, { "epoch": 1.6527159394479074, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.731380462646484, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8781954050064087, "num_tokens": 495524543.0, "step": 12992 }, { "epoch": 1.652843149726498, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.845624923706055, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8609057664871216, "num_tokens": 495562068.0, "step": 12993 }, { "epoch": 1.6529703600050885, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.713837623596191, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8779898881912231, "num_tokens": 495598355.0, "step": 12994 }, { "epoch": 1.6530975702836788, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.787396430969238, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8566974997520447, "num_tokens": 495634735.0, "step": 12995 }, { "epoch": 1.6532247805622693, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.839333534240723, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8698383569717407, "num_tokens": 495668804.0, "step": 12996 }, { "epoch": 1.6533519908408598, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.695924758911133, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8742422461509705, "num_tokens": 495707129.0, "step": 12997 }, { "epoch": 1.6534792011194503, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.784006118774414, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8487718105316162, "num_tokens": 495746501.0, "step": 12998 }, { "epoch": 1.6536064113980409, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.805068016052246, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.885525107383728, "num_tokens": 495789723.0, "step": 12999 }, { "epoch": 1.6537336216766314, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.708442687988281, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8610738515853882, "num_tokens": 495827073.0, "step": 13000 }, { "epoch": 1.653860831955222, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.82831859588623, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8591008186340332, "num_tokens": 495862362.0, "step": 13001 }, { "epoch": 1.6539880422338125, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.709312438964844, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8719420433044434, "num_tokens": 495905546.0, "step": 13002 }, { "epoch": 1.654115252512403, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.820455551147461, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8671784400939941, "num_tokens": 495942291.0, "step": 13003 }, { "epoch": 1.6542424627909935, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.778094291687012, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8594860434532166, "num_tokens": 495978631.0, "step": 13004 }, { "epoch": 1.654369673069584, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.737363815307617, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8644426465034485, "num_tokens": 496016626.0, "step": 13005 }, { "epoch": 1.6544968833481746, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.799947738647461, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.862371563911438, "num_tokens": 496051949.0, "step": 13006 }, { "epoch": 1.654624093626765, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.815372467041016, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.875740647315979, "num_tokens": 496092693.0, "step": 13007 }, { "epoch": 1.6547513039053556, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.760822296142578, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8636517524719238, "num_tokens": 496132853.0, "step": 13008 }, { "epoch": 1.6548785141839462, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.841877937316895, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8551796674728394, "num_tokens": 496168303.0, "step": 13009 }, { "epoch": 1.6550057244625367, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.706707954406738, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8571438789367676, "num_tokens": 496206132.0, "step": 13010 }, { "epoch": 1.6551329347411272, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.841020584106445, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8629868030548096, "num_tokens": 496247542.0, "step": 13011 }, { "epoch": 1.6552601450197177, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.791773796081543, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8788049221038818, "num_tokens": 496285698.0, "step": 13012 }, { "epoch": 1.655387355298308, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.790558815002441, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8715311288833618, "num_tokens": 496322216.0, "step": 13013 }, { "epoch": 1.6555145655768986, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.888387680053711, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8667495846748352, "num_tokens": 496357687.0, "step": 13014 }, { "epoch": 1.655641775855489, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.758944511413574, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8805104494094849, "num_tokens": 496400875.0, "step": 13015 }, { "epoch": 1.6557689861340796, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.860992431640625, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.872483491897583, "num_tokens": 496432076.0, "step": 13016 }, { "epoch": 1.6558961964126702, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.8250150680542, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8670597076416016, "num_tokens": 496466852.0, "step": 13017 }, { "epoch": 1.6560234066912607, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.79931640625, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8681310415267944, "num_tokens": 496508890.0, "step": 13018 }, { "epoch": 1.6561506169698512, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.83105182647705, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8856813907623291, "num_tokens": 496537964.0, "step": 13019 }, { "epoch": 1.6562778272484415, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.864346504211426, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8533886671066284, "num_tokens": 496571263.0, "step": 13020 }, { "epoch": 1.656405037527032, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.778782844543457, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8774933815002441, "num_tokens": 496610058.0, "step": 13021 }, { "epoch": 1.6565322478056226, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.855256080627441, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8628076314926147, "num_tokens": 496640542.0, "step": 13022 }, { "epoch": 1.656659458084213, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.782388687133789, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8676698207855225, "num_tokens": 496678030.0, "step": 13023 }, { "epoch": 1.6567866683628036, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.878117561340332, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8653470873832703, "num_tokens": 496714057.0, "step": 13024 }, { "epoch": 1.6569138786413942, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.79786205291748, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.884300947189331, "num_tokens": 496744501.0, "step": 13025 }, { "epoch": 1.6570410889199847, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.814558029174805, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8743375539779663, "num_tokens": 496780233.0, "step": 13026 }, { "epoch": 1.6571682991985752, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.821871757507324, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8710137605667114, "num_tokens": 496809645.0, "step": 13027 }, { "epoch": 1.6572955094771658, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.830633163452148, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8695908784866333, "num_tokens": 496854871.0, "step": 13028 }, { "epoch": 1.6574227197557563, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.78586483001709, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8852925300598145, "num_tokens": 496891958.0, "step": 13029 }, { "epoch": 1.6575499300343468, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.826452255249023, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8625301718711853, "num_tokens": 496928550.0, "step": 13030 }, { "epoch": 1.6576771403129373, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.775238990783691, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8684291243553162, "num_tokens": 496972009.0, "step": 13031 }, { "epoch": 1.6578043505915279, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.796823501586914, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8682520985603333, "num_tokens": 497013605.0, "step": 13032 }, { "epoch": 1.6579315608701184, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.754505157470703, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8707544207572937, "num_tokens": 497048712.0, "step": 13033 }, { "epoch": 1.658058771148709, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.762280464172363, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8836276531219482, "num_tokens": 497089161.0, "step": 13034 }, { "epoch": 1.6581859814272994, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.813490867614746, "learning_rate": 1e-06, "loss": 0.4929, "mean_token_accuracy": 0.8412594199180603, "num_tokens": 497127253.0, "step": 13035 }, { "epoch": 1.65831319170589, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.749343872070312, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8661206960678101, "num_tokens": 497167637.0, "step": 13036 }, { "epoch": 1.6584404019844805, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.803583145141602, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8698855638504028, "num_tokens": 497205663.0, "step": 13037 }, { "epoch": 1.6585676122630708, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.89404296875, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8860890865325928, "num_tokens": 497242146.0, "step": 13038 }, { "epoch": 1.6586948225416613, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.773530006408691, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8740497827529907, "num_tokens": 497279339.0, "step": 13039 }, { "epoch": 1.6588220328202519, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.771069526672363, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8615787029266357, "num_tokens": 497318530.0, "step": 13040 }, { "epoch": 1.6589492430988424, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.801868438720703, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8638707995414734, "num_tokens": 497352943.0, "step": 13041 }, { "epoch": 1.659076453377433, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.822481155395508, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8635187149047852, "num_tokens": 497387729.0, "step": 13042 }, { "epoch": 1.6592036636560235, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.747021675109863, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8704164028167725, "num_tokens": 497427875.0, "step": 13043 }, { "epoch": 1.6593308739346138, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.793326377868652, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8622658252716064, "num_tokens": 497469076.0, "step": 13044 }, { "epoch": 1.6594580842132043, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.813558578491211, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8711904287338257, "num_tokens": 497509556.0, "step": 13045 }, { "epoch": 1.6595852944917948, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.816880226135254, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8617147207260132, "num_tokens": 497548151.0, "step": 13046 }, { "epoch": 1.6597125047703853, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.787114143371582, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.8434553146362305, "num_tokens": 497589088.0, "step": 13047 }, { "epoch": 1.6598397150489759, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.747475624084473, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8787227272987366, "num_tokens": 497632095.0, "step": 13048 }, { "epoch": 1.6599669253275664, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.8404541015625, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8805901408195496, "num_tokens": 497665148.0, "step": 13049 }, { "epoch": 1.660094135606157, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.80889892578125, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8637810349464417, "num_tokens": 497700134.0, "step": 13050 }, { "epoch": 1.6602213458847475, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.852559089660645, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8744567632675171, "num_tokens": 497743101.0, "step": 13051 }, { "epoch": 1.660348556163338, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.74050235748291, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8631844520568848, "num_tokens": 497775112.0, "step": 13052 }, { "epoch": 1.6604757664419285, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.856782913208008, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8800656199455261, "num_tokens": 497813320.0, "step": 13053 }, { "epoch": 1.660602976720519, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.788315773010254, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8650484085083008, "num_tokens": 497852780.0, "step": 13054 }, { "epoch": 1.6607301869991096, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.826140403747559, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.871687650680542, "num_tokens": 497886313.0, "step": 13055 }, { "epoch": 1.6608573972777, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.82498836517334, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8668056726455688, "num_tokens": 497924108.0, "step": 13056 }, { "epoch": 1.6609846075562906, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.810431480407715, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8593846559524536, "num_tokens": 497964656.0, "step": 13057 }, { "epoch": 1.6611118178348812, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.778609275817871, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.855689525604248, "num_tokens": 498005436.0, "step": 13058 }, { "epoch": 1.6612390281134717, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.808635711669922, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8645332455635071, "num_tokens": 498042237.0, "step": 13059 }, { "epoch": 1.6613662383920622, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.78088092803955, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8683025240898132, "num_tokens": 498083008.0, "step": 13060 }, { "epoch": 1.6614934486706527, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.823247909545898, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8461225032806396, "num_tokens": 498120550.0, "step": 13061 }, { "epoch": 1.661620658949243, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.8335599899292, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8664395809173584, "num_tokens": 498153033.0, "step": 13062 }, { "epoch": 1.6617478692278336, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.75180435180664, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8715252876281738, "num_tokens": 498190491.0, "step": 13063 }, { "epoch": 1.661875079506424, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.751602172851562, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8626103401184082, "num_tokens": 498235889.0, "step": 13064 }, { "epoch": 1.6620022897850146, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.760686874389648, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8661737442016602, "num_tokens": 498277191.0, "step": 13065 }, { "epoch": 1.6621295000636052, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.835700035095215, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8586298227310181, "num_tokens": 498318443.0, "step": 13066 }, { "epoch": 1.6622567103421957, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.78260612487793, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8576542735099792, "num_tokens": 498355955.0, "step": 13067 }, { "epoch": 1.662383920620786, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.831297874450684, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8704351186752319, "num_tokens": 498388588.0, "step": 13068 }, { "epoch": 1.6625111308993765, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.806294441223145, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.867713212966919, "num_tokens": 498430661.0, "step": 13069 }, { "epoch": 1.662638341177967, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.836106300354004, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8569175004959106, "num_tokens": 498470428.0, "step": 13070 }, { "epoch": 1.6627655514565576, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.80027961730957, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8685168027877808, "num_tokens": 498512031.0, "step": 13071 }, { "epoch": 1.662892761735148, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.823625564575195, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8911058902740479, "num_tokens": 498548344.0, "step": 13072 }, { "epoch": 1.6630199720137386, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.778107643127441, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8819288015365601, "num_tokens": 498590115.0, "step": 13073 }, { "epoch": 1.6631471822923292, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.797094345092773, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8699360489845276, "num_tokens": 498634034.0, "step": 13074 }, { "epoch": 1.6632743925709197, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.772204399108887, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8615872263908386, "num_tokens": 498675086.0, "step": 13075 }, { "epoch": 1.6634016028495102, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.842231750488281, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.875129759311676, "num_tokens": 498711374.0, "step": 13076 }, { "epoch": 1.6635288131281007, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.775313377380371, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8741295337677002, "num_tokens": 498753050.0, "step": 13077 }, { "epoch": 1.6636560234066913, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.858304977416992, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8702993392944336, "num_tokens": 498790406.0, "step": 13078 }, { "epoch": 1.6637832336852818, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.80324935913086, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8687958717346191, "num_tokens": 498827712.0, "step": 13079 }, { "epoch": 1.6639104439638723, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.873671531677246, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8738889098167419, "num_tokens": 498862283.0, "step": 13080 }, { "epoch": 1.6640376542424629, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.82602596282959, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8701748251914978, "num_tokens": 498902686.0, "step": 13081 }, { "epoch": 1.6641648645210534, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.847681999206543, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8584188222885132, "num_tokens": 498942785.0, "step": 13082 }, { "epoch": 1.664292074799644, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.852249145507812, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8736156225204468, "num_tokens": 498982761.0, "step": 13083 }, { "epoch": 1.6644192850782344, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.825637817382812, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8505155444145203, "num_tokens": 499025762.0, "step": 13084 }, { "epoch": 1.664546495356825, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.812874794006348, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8604811429977417, "num_tokens": 499063394.0, "step": 13085 }, { "epoch": 1.6646737056354155, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.775562286376953, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8788104057312012, "num_tokens": 499105824.0, "step": 13086 }, { "epoch": 1.6648009159140058, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.857589721679688, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8773725032806396, "num_tokens": 499140935.0, "step": 13087 }, { "epoch": 1.6649281261925963, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.809720039367676, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8683593273162842, "num_tokens": 499179492.0, "step": 13088 }, { "epoch": 1.6650553364711869, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.842706680297852, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.879560112953186, "num_tokens": 499214716.0, "step": 13089 }, { "epoch": 1.6651825467497774, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.809185981750488, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8765734434127808, "num_tokens": 499250337.0, "step": 13090 }, { "epoch": 1.665309757028368, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.899210929870605, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8735243082046509, "num_tokens": 499283675.0, "step": 13091 }, { "epoch": 1.6654369673069584, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.797921180725098, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8736441135406494, "num_tokens": 499319153.0, "step": 13092 }, { "epoch": 1.6655641775855488, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.846545219421387, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.865486741065979, "num_tokens": 499356675.0, "step": 13093 }, { "epoch": 1.6656913878641393, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.81466007232666, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.883840799331665, "num_tokens": 499392129.0, "step": 13094 }, { "epoch": 1.6658185981427298, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.87824535369873, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8806930780410767, "num_tokens": 499424958.0, "step": 13095 }, { "epoch": 1.6659458084213203, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.836215019226074, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8774505257606506, "num_tokens": 499467769.0, "step": 13096 }, { "epoch": 1.6660730186999109, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.797916412353516, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8604873418807983, "num_tokens": 499513828.0, "step": 13097 }, { "epoch": 1.6662002289785014, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.913442611694336, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8747525215148926, "num_tokens": 499548219.0, "step": 13098 }, { "epoch": 1.666327439257092, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.82392692565918, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8526625633239746, "num_tokens": 499587158.0, "step": 13099 }, { "epoch": 1.6664546495356825, "ewc_loss": 0.0302734375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.866867065429688, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8733774423599243, "num_tokens": 499625235.0, "step": 13100 }, { "epoch": 1.666581859814273, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.878910064697266, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8506507873535156, "num_tokens": 499662842.0, "step": 13101 }, { "epoch": 1.6667090700928635, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.824522972106934, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8455730676651001, "num_tokens": 499703464.0, "step": 13102 }, { "epoch": 1.666836280371454, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.858342170715332, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8679944276809692, "num_tokens": 499739160.0, "step": 13103 }, { "epoch": 1.6669634906500446, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.853554725646973, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.855797529220581, "num_tokens": 499775024.0, "step": 13104 }, { "epoch": 1.667090700928635, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.839269638061523, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8726726770401001, "num_tokens": 499816479.0, "step": 13105 }, { "epoch": 1.6672179112072256, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.849930763244629, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8702444434165955, "num_tokens": 499848790.0, "step": 13106 }, { "epoch": 1.6673451214858162, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.754074096679688, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8889334201812744, "num_tokens": 499890955.0, "step": 13107 }, { "epoch": 1.6674723317644067, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.737386703491211, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8783389329910278, "num_tokens": 499924813.0, "step": 13108 }, { "epoch": 1.6675995420429972, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.813817977905273, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8660024404525757, "num_tokens": 499960032.0, "step": 13109 }, { "epoch": 1.6677267523215877, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.732096672058105, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8830915093421936, "num_tokens": 499995868.0, "step": 13110 }, { "epoch": 1.667853962600178, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.867445945739746, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8738344311714172, "num_tokens": 500031662.0, "step": 13111 }, { "epoch": 1.6679811728787686, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.723246574401855, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8601785898208618, "num_tokens": 500070189.0, "step": 13112 }, { "epoch": 1.668108383157359, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.820624351501465, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8661931753158569, "num_tokens": 500109999.0, "step": 13113 }, { "epoch": 1.6682355934359496, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.751094818115234, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8530377149581909, "num_tokens": 500151473.0, "step": 13114 }, { "epoch": 1.6683628037145402, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.757411003112793, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8683760762214661, "num_tokens": 500187436.0, "step": 13115 }, { "epoch": 1.6684900139931307, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.838555335998535, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8691779971122742, "num_tokens": 500222433.0, "step": 13116 }, { "epoch": 1.668617224271721, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.759845733642578, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8693593740463257, "num_tokens": 500257390.0, "step": 13117 }, { "epoch": 1.6687444345503115, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.79636287689209, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8688185214996338, "num_tokens": 500301260.0, "step": 13118 }, { "epoch": 1.668871644828902, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.741236686706543, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8745428323745728, "num_tokens": 500343453.0, "step": 13119 }, { "epoch": 1.6689988551074926, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.822120666503906, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8766409754753113, "num_tokens": 500375133.0, "step": 13120 }, { "epoch": 1.669126065386083, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.8592529296875, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8661144971847534, "num_tokens": 500408540.0, "step": 13121 }, { "epoch": 1.6692532756646736, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.756109237670898, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8823720216751099, "num_tokens": 500446243.0, "step": 13122 }, { "epoch": 1.6693804859432642, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.824249267578125, "learning_rate": 1e-06, "loss": 0.4946, "mean_token_accuracy": 0.8436949253082275, "num_tokens": 500484313.0, "step": 13123 }, { "epoch": 1.6695076962218547, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.76695442199707, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8818035125732422, "num_tokens": 500525566.0, "step": 13124 }, { "epoch": 1.6696349065004452, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.802443504333496, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8753415942192078, "num_tokens": 500566274.0, "step": 13125 }, { "epoch": 1.6697621167790357, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.81064224243164, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8628309369087219, "num_tokens": 500602769.0, "step": 13126 }, { "epoch": 1.6698893270576263, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.840301513671875, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8531253337860107, "num_tokens": 500642895.0, "step": 13127 }, { "epoch": 1.6700165373362168, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.727364540100098, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8673498630523682, "num_tokens": 500681461.0, "step": 13128 }, { "epoch": 1.6701437476148073, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.782469749450684, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8530029058456421, "num_tokens": 500723395.0, "step": 13129 }, { "epoch": 1.6702709578933979, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.796993255615234, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.870825469493866, "num_tokens": 500760571.0, "step": 13130 }, { "epoch": 1.6703981681719884, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.80787181854248, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8609572649002075, "num_tokens": 500797858.0, "step": 13131 }, { "epoch": 1.670525378450579, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.80215835571289, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8758345246315002, "num_tokens": 500840199.0, "step": 13132 }, { "epoch": 1.6706525887291694, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.779790878295898, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8633497953414917, "num_tokens": 500883290.0, "step": 13133 }, { "epoch": 1.67077979900776, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.802682876586914, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8727813363075256, "num_tokens": 500923072.0, "step": 13134 }, { "epoch": 1.6709070092863505, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.787324905395508, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.87322998046875, "num_tokens": 500958726.0, "step": 13135 }, { "epoch": 1.6710342195649408, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.82756233215332, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8663076162338257, "num_tokens": 500996746.0, "step": 13136 }, { "epoch": 1.6711614298435313, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.899190902709961, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8619771003723145, "num_tokens": 501034135.0, "step": 13137 }, { "epoch": 1.6712886401221219, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.718367576599121, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.872107744216919, "num_tokens": 501069810.0, "step": 13138 }, { "epoch": 1.6714158504007124, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.812413215637207, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.862877368927002, "num_tokens": 501110597.0, "step": 13139 }, { "epoch": 1.671543060679303, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.870655059814453, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8591340780258179, "num_tokens": 501145927.0, "step": 13140 }, { "epoch": 1.6716702709578934, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.794928550720215, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8506108522415161, "num_tokens": 501188473.0, "step": 13141 }, { "epoch": 1.6717974812364838, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.82888412475586, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8813673853874207, "num_tokens": 501226143.0, "step": 13142 }, { "epoch": 1.6719246915150743, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.810354232788086, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8605438470840454, "num_tokens": 501266980.0, "step": 13143 }, { "epoch": 1.6720519017936648, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.771876335144043, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8658062815666199, "num_tokens": 501305414.0, "step": 13144 }, { "epoch": 1.6721791120722553, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.830451965332031, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8611671328544617, "num_tokens": 501343009.0, "step": 13145 }, { "epoch": 1.6723063223508459, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.873245239257812, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8748809099197388, "num_tokens": 501383571.0, "step": 13146 }, { "epoch": 1.6724335326294364, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.788056373596191, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8699284195899963, "num_tokens": 501425444.0, "step": 13147 }, { "epoch": 1.672560742908027, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.811976432800293, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8817625641822815, "num_tokens": 501466536.0, "step": 13148 }, { "epoch": 1.6726879531866174, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.9055814743042, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8709089756011963, "num_tokens": 501507346.0, "step": 13149 }, { "epoch": 1.672815163465208, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.779415130615234, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8758659362792969, "num_tokens": 501545515.0, "step": 13150 }, { "epoch": 1.6729423737437985, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.856100082397461, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8629099130630493, "num_tokens": 501588793.0, "step": 13151 }, { "epoch": 1.673069584022389, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.799087524414062, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8660691380500793, "num_tokens": 501624384.0, "step": 13152 }, { "epoch": 1.6731967943009796, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.81738567352295, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8686685562133789, "num_tokens": 501663150.0, "step": 13153 }, { "epoch": 1.67332400457957, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.786659240722656, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8621779680252075, "num_tokens": 501706177.0, "step": 13154 }, { "epoch": 1.6734512148581606, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.803483963012695, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8808153867721558, "num_tokens": 501745267.0, "step": 13155 }, { "epoch": 1.6735784251367511, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.845965385437012, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8645653128623962, "num_tokens": 501783591.0, "step": 13156 }, { "epoch": 1.6737056354153417, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.87867259979248, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8600345849990845, "num_tokens": 501825190.0, "step": 13157 }, { "epoch": 1.6738328456939322, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.848100662231445, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8689786195755005, "num_tokens": 501855813.0, "step": 13158 }, { "epoch": 1.6739600559725227, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.888876914978027, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8648901581764221, "num_tokens": 501890106.0, "step": 13159 }, { "epoch": 1.674087266251113, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.786168098449707, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8558017015457153, "num_tokens": 501935437.0, "step": 13160 }, { "epoch": 1.6742144765297036, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.893378257751465, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8556370735168457, "num_tokens": 501980134.0, "step": 13161 }, { "epoch": 1.674341686808294, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.76855754852295, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8625149726867676, "num_tokens": 502015977.0, "step": 13162 }, { "epoch": 1.6744688970868846, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.787124633789062, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8718825578689575, "num_tokens": 502061717.0, "step": 13163 }, { "epoch": 1.6745961073654752, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.8451509475708, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8662843704223633, "num_tokens": 502101594.0, "step": 13164 }, { "epoch": 1.6747233176440657, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.860503196716309, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8681997060775757, "num_tokens": 502136521.0, "step": 13165 }, { "epoch": 1.674850527922656, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.824663162231445, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8502201437950134, "num_tokens": 502174821.0, "step": 13166 }, { "epoch": 1.6749777382012465, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.798534393310547, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8626094460487366, "num_tokens": 502210615.0, "step": 13167 }, { "epoch": 1.675104948479837, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.824953079223633, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8689151406288147, "num_tokens": 502248220.0, "step": 13168 }, { "epoch": 1.6752321587584276, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.765644073486328, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8673678636550903, "num_tokens": 502290686.0, "step": 13169 }, { "epoch": 1.675359369037018, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.85215950012207, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8568621873855591, "num_tokens": 502331789.0, "step": 13170 }, { "epoch": 1.6754865793156086, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.696907043457031, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8760780096054077, "num_tokens": 502371490.0, "step": 13171 }, { "epoch": 1.6756137895941992, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.833352088928223, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8673584461212158, "num_tokens": 502415606.0, "step": 13172 }, { "epoch": 1.6757409998727897, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.80094051361084, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8673706650733948, "num_tokens": 502456203.0, "step": 13173 }, { "epoch": 1.6758682101513802, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.863369941711426, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8846511244773865, "num_tokens": 502491508.0, "step": 13174 }, { "epoch": 1.6759954204299707, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.827921867370605, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8575023412704468, "num_tokens": 502533408.0, "step": 13175 }, { "epoch": 1.6761226307085613, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.811483383178711, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8615906834602356, "num_tokens": 502569824.0, "step": 13176 }, { "epoch": 1.6762498409871518, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.831193923950195, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8761396408081055, "num_tokens": 502604776.0, "step": 13177 }, { "epoch": 1.6763770512657423, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.814456939697266, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8510100841522217, "num_tokens": 502645732.0, "step": 13178 }, { "epoch": 1.6765042615443329, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.817745208740234, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.871894121170044, "num_tokens": 502686044.0, "step": 13179 }, { "epoch": 1.6766314718229234, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.763227462768555, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8589349985122681, "num_tokens": 502728288.0, "step": 13180 }, { "epoch": 1.676758682101514, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.798463821411133, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.860412061214447, "num_tokens": 502760772.0, "step": 13181 }, { "epoch": 1.6768858923801044, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.817476272583008, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8683055639266968, "num_tokens": 502791585.0, "step": 13182 }, { "epoch": 1.677013102658695, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.805495262145996, "learning_rate": 1e-06, "loss": 0.3601, "mean_token_accuracy": 0.8852672576904297, "num_tokens": 502826339.0, "step": 13183 }, { "epoch": 1.6771403129372855, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.790189743041992, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8845130801200867, "num_tokens": 502868716.0, "step": 13184 }, { "epoch": 1.6772675232158758, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.819884300231934, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8686860203742981, "num_tokens": 502907992.0, "step": 13185 }, { "epoch": 1.6773947334944663, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.809680938720703, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8629450798034668, "num_tokens": 502947197.0, "step": 13186 }, { "epoch": 1.6775219437730569, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.770500183105469, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8773280382156372, "num_tokens": 502985890.0, "step": 13187 }, { "epoch": 1.6776491540516474, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.794381141662598, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8634157180786133, "num_tokens": 503033476.0, "step": 13188 }, { "epoch": 1.677776364330238, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.823232650756836, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8766522407531738, "num_tokens": 503068002.0, "step": 13189 }, { "epoch": 1.6779035746088284, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.85427474975586, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8615455031394958, "num_tokens": 503110059.0, "step": 13190 }, { "epoch": 1.6780307848874187, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.808768272399902, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8614481687545776, "num_tokens": 503146617.0, "step": 13191 }, { "epoch": 1.6781579951660093, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.830561637878418, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8660815358161926, "num_tokens": 503186436.0, "step": 13192 }, { "epoch": 1.6782852054445998, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.864603042602539, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8743116855621338, "num_tokens": 503227983.0, "step": 13193 }, { "epoch": 1.6784124157231903, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.808279037475586, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8553264141082764, "num_tokens": 503269301.0, "step": 13194 }, { "epoch": 1.6785396260017809, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.865413665771484, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8510879278182983, "num_tokens": 503308919.0, "step": 13195 }, { "epoch": 1.6786668362803714, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.791829109191895, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8777557611465454, "num_tokens": 503345030.0, "step": 13196 }, { "epoch": 1.678794046558962, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.850959777832031, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8602316975593567, "num_tokens": 503379142.0, "step": 13197 }, { "epoch": 1.6789212568375524, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.840229988098145, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8678358793258667, "num_tokens": 503411810.0, "step": 13198 }, { "epoch": 1.679048467116143, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.765628814697266, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8663289546966553, "num_tokens": 503450266.0, "step": 13199 }, { "epoch": 1.6791756773947335, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.824078559875488, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8757184743881226, "num_tokens": 503491015.0, "step": 13200 }, { "epoch": 1.679302887673324, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.80081558227539, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8765648603439331, "num_tokens": 503531964.0, "step": 13201 }, { "epoch": 1.6794300979519146, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.825458526611328, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8725394010543823, "num_tokens": 503576069.0, "step": 13202 }, { "epoch": 1.679557308230505, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.82452392578125, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8689403533935547, "num_tokens": 503607895.0, "step": 13203 }, { "epoch": 1.6796845185090956, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.856624603271484, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8625656366348267, "num_tokens": 503636891.0, "step": 13204 }, { "epoch": 1.6798117287876861, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.832733154296875, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8819912672042847, "num_tokens": 503682037.0, "step": 13205 }, { "epoch": 1.6799389390662767, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.799765586853027, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.871558666229248, "num_tokens": 503724008.0, "step": 13206 }, { "epoch": 1.6800661493448672, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.857536315917969, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.876584529876709, "num_tokens": 503759008.0, "step": 13207 }, { "epoch": 1.6801933596234577, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.805708885192871, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8648837208747864, "num_tokens": 503796599.0, "step": 13208 }, { "epoch": 1.680320569902048, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.796335220336914, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8656339645385742, "num_tokens": 503835539.0, "step": 13209 }, { "epoch": 1.6804477801806386, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.807977676391602, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8653900623321533, "num_tokens": 503875933.0, "step": 13210 }, { "epoch": 1.680574990459229, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.844538688659668, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8654173612594604, "num_tokens": 503913740.0, "step": 13211 }, { "epoch": 1.6807022007378196, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.813128471374512, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.871921956539154, "num_tokens": 503955105.0, "step": 13212 }, { "epoch": 1.6808294110164101, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.906898498535156, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8679094910621643, "num_tokens": 503987641.0, "step": 13213 }, { "epoch": 1.6809566212950007, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.884523391723633, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8596047163009644, "num_tokens": 504028699.0, "step": 13214 }, { "epoch": 1.681083831573591, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.828935623168945, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8617749214172363, "num_tokens": 504065597.0, "step": 13215 }, { "epoch": 1.6812110418521815, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.906250953674316, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8541775941848755, "num_tokens": 504105968.0, "step": 13216 }, { "epoch": 1.681338252130772, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.79621410369873, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8707844614982605, "num_tokens": 504143743.0, "step": 13217 }, { "epoch": 1.6814654624093626, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.849698066711426, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8703993558883667, "num_tokens": 504177023.0, "step": 13218 }, { "epoch": 1.681592672687953, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.912291526794434, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8671859502792358, "num_tokens": 504210492.0, "step": 13219 }, { "epoch": 1.6817198829665436, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.822861671447754, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.868730366230011, "num_tokens": 504248322.0, "step": 13220 }, { "epoch": 1.6818470932451342, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.82113265991211, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8544343709945679, "num_tokens": 504285915.0, "step": 13221 }, { "epoch": 1.6819743035237247, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.874299049377441, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8664419651031494, "num_tokens": 504325578.0, "step": 13222 }, { "epoch": 1.6821015138023152, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.895377159118652, "learning_rate": 1e-06, "loss": 0.3387, "mean_token_accuracy": 0.8878981471061707, "num_tokens": 504362876.0, "step": 13223 }, { "epoch": 1.6822287240809057, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.910430908203125, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8667818307876587, "num_tokens": 504402303.0, "step": 13224 }, { "epoch": 1.6823559343594963, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.878312110900879, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8695223331451416, "num_tokens": 504440039.0, "step": 13225 }, { "epoch": 1.6824831446380868, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.867841720581055, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8596790432929993, "num_tokens": 504485102.0, "step": 13226 }, { "epoch": 1.6826103549166773, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.831899642944336, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8665724992752075, "num_tokens": 504522762.0, "step": 13227 }, { "epoch": 1.6827375651952678, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.91191291809082, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8723318576812744, "num_tokens": 504562818.0, "step": 13228 }, { "epoch": 1.6828647754738584, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.854484558105469, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8591129779815674, "num_tokens": 504602914.0, "step": 13229 }, { "epoch": 1.682991985752449, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.880274772644043, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.875346839427948, "num_tokens": 504641376.0, "step": 13230 }, { "epoch": 1.6831191960310394, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.859413146972656, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8746318817138672, "num_tokens": 504675031.0, "step": 13231 }, { "epoch": 1.68324640630963, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.915532112121582, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8753256797790527, "num_tokens": 504709624.0, "step": 13232 }, { "epoch": 1.6833736165882205, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.86196231842041, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8724000453948975, "num_tokens": 504749310.0, "step": 13233 }, { "epoch": 1.6835008268668108, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.886678695678711, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8688799738883972, "num_tokens": 504791700.0, "step": 13234 }, { "epoch": 1.6836280371454013, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.889283180236816, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8582170605659485, "num_tokens": 504833246.0, "step": 13235 }, { "epoch": 1.6837552474239919, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.83841609954834, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8729894757270813, "num_tokens": 504872901.0, "step": 13236 }, { "epoch": 1.6838824577025824, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.958199501037598, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8625106811523438, "num_tokens": 504906138.0, "step": 13237 }, { "epoch": 1.684009667981173, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.78663444519043, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8489457368850708, "num_tokens": 504947147.0, "step": 13238 }, { "epoch": 1.6841368782597634, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.944921493530273, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8697699904441833, "num_tokens": 504985332.0, "step": 13239 }, { "epoch": 1.6842640885383537, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.857416152954102, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8593254089355469, "num_tokens": 505027995.0, "step": 13240 }, { "epoch": 1.6843912988169443, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.88403606414795, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8709304332733154, "num_tokens": 505065311.0, "step": 13241 }, { "epoch": 1.6845185090955348, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.852924346923828, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8641834259033203, "num_tokens": 505105810.0, "step": 13242 }, { "epoch": 1.6846457193741253, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.845776557922363, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8746949434280396, "num_tokens": 505141003.0, "step": 13243 }, { "epoch": 1.6847729296527159, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.852486610412598, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8582323789596558, "num_tokens": 505188025.0, "step": 13244 }, { "epoch": 1.6849001399313064, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.826828956604004, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8827534914016724, "num_tokens": 505226101.0, "step": 13245 }, { "epoch": 1.685027350209897, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.858647346496582, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8584995865821838, "num_tokens": 505264880.0, "step": 13246 }, { "epoch": 1.6851545604884874, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.86988353729248, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8624386787414551, "num_tokens": 505306208.0, "step": 13247 }, { "epoch": 1.685281770767078, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.893564224243164, "learning_rate": 1e-06, "loss": 0.3455, "mean_token_accuracy": 0.8891128301620483, "num_tokens": 505346024.0, "step": 13248 }, { "epoch": 1.6854089810456685, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.877771377563477, "learning_rate": 1e-06, "loss": 0.3411, "mean_token_accuracy": 0.8905489444732666, "num_tokens": 505383644.0, "step": 13249 }, { "epoch": 1.685536191324259, "ewc_loss": 0.0302734375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.79107666015625, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8725433945655823, "num_tokens": 505420365.0, "step": 13250 }, { "epoch": 1.6856634016028496, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.892873764038086, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.870585024356842, "num_tokens": 505458211.0, "step": 13251 }, { "epoch": 1.68579061188144, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.866605758666992, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8637147545814514, "num_tokens": 505501649.0, "step": 13252 }, { "epoch": 1.6859178221600306, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.812697410583496, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8711283802986145, "num_tokens": 505538763.0, "step": 13253 }, { "epoch": 1.6860450324386211, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.89246654510498, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.8907127380371094, "num_tokens": 505574562.0, "step": 13254 }, { "epoch": 1.6861722427172117, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.829324722290039, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8823326826095581, "num_tokens": 505609718.0, "step": 13255 }, { "epoch": 1.6862994529958022, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.89810848236084, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.881493091583252, "num_tokens": 505644337.0, "step": 13256 }, { "epoch": 1.6864266632743927, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.825032234191895, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8788903951644897, "num_tokens": 505681929.0, "step": 13257 }, { "epoch": 1.686553873552983, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.830124855041504, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8465968370437622, "num_tokens": 505719373.0, "step": 13258 }, { "epoch": 1.6866810838315736, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.914971351623535, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8552066087722778, "num_tokens": 505757767.0, "step": 13259 }, { "epoch": 1.686808294110164, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.79715347290039, "learning_rate": 1e-06, "loss": 0.5378, "mean_token_accuracy": 0.8378877639770508, "num_tokens": 505794777.0, "step": 13260 }, { "epoch": 1.6869355043887546, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.892661094665527, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8758999109268188, "num_tokens": 505832672.0, "step": 13261 }, { "epoch": 1.6870627146673451, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.935336112976074, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.857421875, "num_tokens": 505873142.0, "step": 13262 }, { "epoch": 1.6871899249459357, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.823185920715332, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8715496063232422, "num_tokens": 505911994.0, "step": 13263 }, { "epoch": 1.687317135224526, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.87621021270752, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8721585869789124, "num_tokens": 505948874.0, "step": 13264 }, { "epoch": 1.6874443455031165, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.853887557983398, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8746278285980225, "num_tokens": 505989679.0, "step": 13265 }, { "epoch": 1.687571555781707, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.80884075164795, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8774493336677551, "num_tokens": 506033209.0, "step": 13266 }, { "epoch": 1.6876987660602976, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.874068260192871, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8778525590896606, "num_tokens": 506065379.0, "step": 13267 }, { "epoch": 1.687825976338888, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.831297874450684, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8602920770645142, "num_tokens": 506099484.0, "step": 13268 }, { "epoch": 1.6879531866174786, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.875996589660645, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8525574207305908, "num_tokens": 506136559.0, "step": 13269 }, { "epoch": 1.6880803968960691, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.894065856933594, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8676287531852722, "num_tokens": 506172997.0, "step": 13270 }, { "epoch": 1.6882076071746597, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.837957382202148, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8830781579017639, "num_tokens": 506209858.0, "step": 13271 }, { "epoch": 1.6883348174532502, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.821197509765625, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.881414532661438, "num_tokens": 506242260.0, "step": 13272 }, { "epoch": 1.6884620277318407, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.86020565032959, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8569046258926392, "num_tokens": 506276856.0, "step": 13273 }, { "epoch": 1.6885892380104313, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.817115783691406, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8869560956954956, "num_tokens": 506311469.0, "step": 13274 }, { "epoch": 1.6887164482890218, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.860733985900879, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8794243931770325, "num_tokens": 506348240.0, "step": 13275 }, { "epoch": 1.6888436585676123, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.80584716796875, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8704249858856201, "num_tokens": 506387203.0, "step": 13276 }, { "epoch": 1.6889708688462028, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.925777435302734, "learning_rate": 1e-06, "loss": 0.4858, "mean_token_accuracy": 0.8467079401016235, "num_tokens": 506421869.0, "step": 13277 }, { "epoch": 1.6890980791247934, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.832087516784668, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.86954665184021, "num_tokens": 506458873.0, "step": 13278 }, { "epoch": 1.689225289403384, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.833564758300781, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8560582399368286, "num_tokens": 506495947.0, "step": 13279 }, { "epoch": 1.6893524996819744, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.9428129196167, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8507417440414429, "num_tokens": 506533098.0, "step": 13280 }, { "epoch": 1.689479709960565, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.885235786437988, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8790304660797119, "num_tokens": 506571346.0, "step": 13281 }, { "epoch": 1.6896069202391555, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.77011489868164, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8710815906524658, "num_tokens": 506607912.0, "step": 13282 }, { "epoch": 1.6897341305177458, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.890101432800293, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8579277992248535, "num_tokens": 506654258.0, "step": 13283 }, { "epoch": 1.6898613407963363, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.765987396240234, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8760383129119873, "num_tokens": 506693436.0, "step": 13284 }, { "epoch": 1.6899885510749268, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.89869213104248, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.871457576751709, "num_tokens": 506737383.0, "step": 13285 }, { "epoch": 1.6901157613535174, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.84889030456543, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8786361217498779, "num_tokens": 506772282.0, "step": 13286 }, { "epoch": 1.690242971632108, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.89375114440918, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8801553249359131, "num_tokens": 506807208.0, "step": 13287 }, { "epoch": 1.6903701819106984, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.855158805847168, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8748562335968018, "num_tokens": 506847952.0, "step": 13288 }, { "epoch": 1.6904973921892887, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.886786460876465, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8725221157073975, "num_tokens": 506888265.0, "step": 13289 }, { "epoch": 1.6906246024678793, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.912986755371094, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8722800612449646, "num_tokens": 506926096.0, "step": 13290 }, { "epoch": 1.6907518127464698, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.79852294921875, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8630994558334351, "num_tokens": 506962252.0, "step": 13291 }, { "epoch": 1.6908790230250603, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.95063304901123, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8753919005393982, "num_tokens": 507004709.0, "step": 13292 }, { "epoch": 1.6910062333036509, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.788005828857422, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8848097324371338, "num_tokens": 507041041.0, "step": 13293 }, { "epoch": 1.6911334435822414, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.954118728637695, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8632687926292419, "num_tokens": 507080251.0, "step": 13294 }, { "epoch": 1.691260653860832, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.859124183654785, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.86893230676651, "num_tokens": 507120813.0, "step": 13295 }, { "epoch": 1.6913878641394224, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.878958702087402, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8559167385101318, "num_tokens": 507160284.0, "step": 13296 }, { "epoch": 1.691515074418013, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.901966094970703, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8619073629379272, "num_tokens": 507205069.0, "step": 13297 }, { "epoch": 1.6916422846966035, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.818238258361816, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8776658773422241, "num_tokens": 507244689.0, "step": 13298 }, { "epoch": 1.691769494975194, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.875775337219238, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8710761666297913, "num_tokens": 507283917.0, "step": 13299 }, { "epoch": 1.6918967052537845, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.879362106323242, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.872160792350769, "num_tokens": 507327937.0, "step": 13300 }, { "epoch": 1.692023915532375, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.833823204040527, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8834700584411621, "num_tokens": 507369521.0, "step": 13301 }, { "epoch": 1.6921511258109656, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.927390098571777, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8797467947006226, "num_tokens": 507408157.0, "step": 13302 }, { "epoch": 1.6922783360895561, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.861224174499512, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8756381869316101, "num_tokens": 507448003.0, "step": 13303 }, { "epoch": 1.6924055463681467, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.87327766418457, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.877244234085083, "num_tokens": 507489970.0, "step": 13304 }, { "epoch": 1.6925327566467372, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.856220245361328, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8716335892677307, "num_tokens": 507524536.0, "step": 13305 }, { "epoch": 1.6926599669253277, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.91108512878418, "learning_rate": 1e-06, "loss": 0.3322, "mean_token_accuracy": 0.8891403675079346, "num_tokens": 507561575.0, "step": 13306 }, { "epoch": 1.692787177203918, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.835797309875488, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8718302249908447, "num_tokens": 507597212.0, "step": 13307 }, { "epoch": 1.6929143874825086, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.886096954345703, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8720051646232605, "num_tokens": 507634251.0, "step": 13308 }, { "epoch": 1.693041597761099, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.840789794921875, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.875870943069458, "num_tokens": 507671345.0, "step": 13309 }, { "epoch": 1.6931688080396896, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.892879486083984, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8592832088470459, "num_tokens": 507709797.0, "step": 13310 }, { "epoch": 1.6932960183182801, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.82127571105957, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8685064911842346, "num_tokens": 507747047.0, "step": 13311 }, { "epoch": 1.6934232285968707, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.879545211791992, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8697081804275513, "num_tokens": 507782187.0, "step": 13312 }, { "epoch": 1.693550438875461, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.88142204284668, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8567101359367371, "num_tokens": 507826714.0, "step": 13313 }, { "epoch": 1.6936776491540515, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.851396560668945, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8611899614334106, "num_tokens": 507869423.0, "step": 13314 }, { "epoch": 1.693804859432642, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.832569122314453, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.879565954208374, "num_tokens": 507904711.0, "step": 13315 }, { "epoch": 1.6939320697112326, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.841958045959473, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.864788830280304, "num_tokens": 507947760.0, "step": 13316 }, { "epoch": 1.694059279989823, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.882818222045898, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8670003414154053, "num_tokens": 507984249.0, "step": 13317 }, { "epoch": 1.6941864902684136, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.88413143157959, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8657496571540833, "num_tokens": 508017206.0, "step": 13318 }, { "epoch": 1.6943137005470041, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.902791023254395, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8649787902832031, "num_tokens": 508056357.0, "step": 13319 }, { "epoch": 1.6944409108255947, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.81659984588623, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.860829770565033, "num_tokens": 508094870.0, "step": 13320 }, { "epoch": 1.6945681211041852, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.929713249206543, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8682293891906738, "num_tokens": 508120323.0, "step": 13321 }, { "epoch": 1.6946953313827757, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.82258415222168, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8611117601394653, "num_tokens": 508160018.0, "step": 13322 }, { "epoch": 1.6948225416613663, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.8582763671875, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8865435719490051, "num_tokens": 508200947.0, "step": 13323 }, { "epoch": 1.6949497519399568, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.7867431640625, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8561305999755859, "num_tokens": 508246457.0, "step": 13324 }, { "epoch": 1.6950769622185473, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.863470077514648, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8648015856742859, "num_tokens": 508281951.0, "step": 13325 }, { "epoch": 1.6952041724971378, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.819037437438965, "learning_rate": 1e-06, "loss": 0.4749, "mean_token_accuracy": 0.8513315916061401, "num_tokens": 508322571.0, "step": 13326 }, { "epoch": 1.6953313827757284, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.798890113830566, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8766807317733765, "num_tokens": 508357347.0, "step": 13327 }, { "epoch": 1.695458593054319, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.842660903930664, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8709183931350708, "num_tokens": 508401813.0, "step": 13328 }, { "epoch": 1.6955858033329094, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.87365436553955, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.857063889503479, "num_tokens": 508440457.0, "step": 13329 }, { "epoch": 1.6957130136115, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.885851860046387, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8597873449325562, "num_tokens": 508472272.0, "step": 13330 }, { "epoch": 1.6958402238900905, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.870565414428711, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8813822269439697, "num_tokens": 508509352.0, "step": 13331 }, { "epoch": 1.6959674341686808, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.835492134094238, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8783042430877686, "num_tokens": 508552515.0, "step": 13332 }, { "epoch": 1.6960946444472713, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.883030891418457, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8821510076522827, "num_tokens": 508591399.0, "step": 13333 }, { "epoch": 1.6962218547258618, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.897293090820312, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8736063241958618, "num_tokens": 508630427.0, "step": 13334 }, { "epoch": 1.6963490650044524, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.869526863098145, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8767942190170288, "num_tokens": 508667469.0, "step": 13335 }, { "epoch": 1.696476275283043, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.884556770324707, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8727349638938904, "num_tokens": 508706308.0, "step": 13336 }, { "epoch": 1.6966034855616334, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.868330001831055, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.857302188873291, "num_tokens": 508738968.0, "step": 13337 }, { "epoch": 1.6967306958402237, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.826604843139648, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8752784729003906, "num_tokens": 508769879.0, "step": 13338 }, { "epoch": 1.6968579061188143, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.80386734008789, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8538076877593994, "num_tokens": 508812277.0, "step": 13339 }, { "epoch": 1.6969851163974048, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.827958106994629, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8590164184570312, "num_tokens": 508849293.0, "step": 13340 }, { "epoch": 1.6971123266759953, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.828797340393066, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8709798455238342, "num_tokens": 508885798.0, "step": 13341 }, { "epoch": 1.6972395369545858, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.82420539855957, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8673980236053467, "num_tokens": 508920491.0, "step": 13342 }, { "epoch": 1.6973667472331764, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.808902740478516, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8691346645355225, "num_tokens": 508956161.0, "step": 13343 }, { "epoch": 1.697493957511767, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.761847496032715, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8653903007507324, "num_tokens": 508995696.0, "step": 13344 }, { "epoch": 1.6976211677903574, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.858918190002441, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8740900754928589, "num_tokens": 509029505.0, "step": 13345 }, { "epoch": 1.697748378068948, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.891109466552734, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8765658736228943, "num_tokens": 509062815.0, "step": 13346 }, { "epoch": 1.6978755883475385, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.787569999694824, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8754798173904419, "num_tokens": 509103376.0, "step": 13347 }, { "epoch": 1.698002798626129, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.781926155090332, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8801252841949463, "num_tokens": 509142797.0, "step": 13348 }, { "epoch": 1.6981300089047195, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.780426979064941, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8586118221282959, "num_tokens": 509182724.0, "step": 13349 }, { "epoch": 1.69825721918331, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.872651100158691, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8780783414840698, "num_tokens": 509217856.0, "step": 13350 }, { "epoch": 1.6983844294619006, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.800923347473145, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8698067665100098, "num_tokens": 509251581.0, "step": 13351 }, { "epoch": 1.6985116397404911, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.825377464294434, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8857120275497437, "num_tokens": 509286687.0, "step": 13352 }, { "epoch": 1.6986388500190817, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.809298515319824, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8645599484443665, "num_tokens": 509323372.0, "step": 13353 }, { "epoch": 1.6987660602976722, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.848360061645508, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8618323802947998, "num_tokens": 509362472.0, "step": 13354 }, { "epoch": 1.6988932705762627, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.83170223236084, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8705223798751831, "num_tokens": 509397676.0, "step": 13355 }, { "epoch": 1.699020480854853, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.81597900390625, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8733748197555542, "num_tokens": 509441100.0, "step": 13356 }, { "epoch": 1.6991476911334435, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.789432525634766, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8652777671813965, "num_tokens": 509485430.0, "step": 13357 }, { "epoch": 1.699274901412034, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.90202808380127, "learning_rate": 1e-06, "loss": 0.506, "mean_token_accuracy": 0.8401677012443542, "num_tokens": 509520102.0, "step": 13358 }, { "epoch": 1.6994021116906246, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.87016773223877, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8711583018302917, "num_tokens": 509556525.0, "step": 13359 }, { "epoch": 1.6995293219692151, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.772970199584961, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8637717366218567, "num_tokens": 509596899.0, "step": 13360 }, { "epoch": 1.6996565322478057, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.885780334472656, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8699315190315247, "num_tokens": 509632125.0, "step": 13361 }, { "epoch": 1.699783742526396, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.785194396972656, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8654364347457886, "num_tokens": 509668131.0, "step": 13362 }, { "epoch": 1.6999109528049865, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.81852912902832, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8816992044448853, "num_tokens": 509704078.0, "step": 13363 }, { "epoch": 1.700038163083577, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.828432083129883, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.882437527179718, "num_tokens": 509739464.0, "step": 13364 }, { "epoch": 1.7001653733621676, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.839790344238281, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8683290481567383, "num_tokens": 509777970.0, "step": 13365 }, { "epoch": 1.700292583640758, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.82305908203125, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8555264472961426, "num_tokens": 509817010.0, "step": 13366 }, { "epoch": 1.7004197939193486, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.866523742675781, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8750612735748291, "num_tokens": 509854192.0, "step": 13367 }, { "epoch": 1.7005470041979391, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.835240364074707, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.883450448513031, "num_tokens": 509888633.0, "step": 13368 }, { "epoch": 1.7006742144765297, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.78571891784668, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8704671859741211, "num_tokens": 509928474.0, "step": 13369 }, { "epoch": 1.7008014247551202, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.854106903076172, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.862127423286438, "num_tokens": 509968875.0, "step": 13370 }, { "epoch": 1.7009286350337107, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.8621826171875, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8705040216445923, "num_tokens": 510006021.0, "step": 13371 }, { "epoch": 1.7010558453123013, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.847914695739746, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8752809762954712, "num_tokens": 510042111.0, "step": 13372 }, { "epoch": 1.7011830555908918, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.82588005065918, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8752245306968689, "num_tokens": 510074894.0, "step": 13373 }, { "epoch": 1.7013102658694823, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.817574501037598, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8614922165870667, "num_tokens": 510111516.0, "step": 13374 }, { "epoch": 1.7014374761480728, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.854387283325195, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8728356957435608, "num_tokens": 510151549.0, "step": 13375 }, { "epoch": 1.7015646864266634, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.866876602172852, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8696940541267395, "num_tokens": 510189215.0, "step": 13376 }, { "epoch": 1.701691896705254, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.77652359008789, "learning_rate": 1e-06, "loss": 0.3474, "mean_token_accuracy": 0.8890652656555176, "num_tokens": 510231189.0, "step": 13377 }, { "epoch": 1.7018191069838444, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.868053436279297, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8732953667640686, "num_tokens": 510273956.0, "step": 13378 }, { "epoch": 1.701946317262435, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.824565887451172, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8857898116111755, "num_tokens": 510315095.0, "step": 13379 }, { "epoch": 1.7020735275410255, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.858972549438477, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8495136499404907, "num_tokens": 510350753.0, "step": 13380 }, { "epoch": 1.7022007378196158, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.891966819763184, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8737661838531494, "num_tokens": 510389257.0, "step": 13381 }, { "epoch": 1.7023279480982063, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.89814567565918, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8711137175559998, "num_tokens": 510428879.0, "step": 13382 }, { "epoch": 1.7024551583767968, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.825440406799316, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8733216524124146, "num_tokens": 510468409.0, "step": 13383 }, { "epoch": 1.7025823686553874, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.840481758117676, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8846338391304016, "num_tokens": 510503988.0, "step": 13384 }, { "epoch": 1.702709578933978, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.908965110778809, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8680160045623779, "num_tokens": 510539009.0, "step": 13385 }, { "epoch": 1.7028367892125684, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.888018608093262, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.884946346282959, "num_tokens": 510576177.0, "step": 13386 }, { "epoch": 1.7029639994911587, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.933820724487305, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8699260950088501, "num_tokens": 510622194.0, "step": 13387 }, { "epoch": 1.7030912097697493, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.808389663696289, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8859083652496338, "num_tokens": 510663564.0, "step": 13388 }, { "epoch": 1.7032184200483398, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.881441116333008, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8803645372390747, "num_tokens": 510702153.0, "step": 13389 }, { "epoch": 1.7033456303269303, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.889411926269531, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8662467002868652, "num_tokens": 510742812.0, "step": 13390 }, { "epoch": 1.7034728406055208, "ewc_loss": 0.030517578125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.80540943145752, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8646501898765564, "num_tokens": 510783363.0, "step": 13391 }, { "epoch": 1.7036000508841114, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.901552200317383, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8560545444488525, "num_tokens": 510818362.0, "step": 13392 }, { "epoch": 1.703727261162702, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.833271026611328, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8818038702011108, "num_tokens": 510854732.0, "step": 13393 }, { "epoch": 1.7038544714412924, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.882986068725586, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8640530705451965, "num_tokens": 510895107.0, "step": 13394 }, { "epoch": 1.703981681719883, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.81747817993164, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8674092292785645, "num_tokens": 510930737.0, "step": 13395 }, { "epoch": 1.7041088919984735, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.866905212402344, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.878034234046936, "num_tokens": 510964646.0, "step": 13396 }, { "epoch": 1.704236102277064, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.86882495880127, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.873371958732605, "num_tokens": 511000583.0, "step": 13397 }, { "epoch": 1.7043633125556545, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.889179229736328, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8728649616241455, "num_tokens": 511033729.0, "step": 13398 }, { "epoch": 1.704490522834245, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.834583282470703, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8750150203704834, "num_tokens": 511074099.0, "step": 13399 }, { "epoch": 1.7046177331128356, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.867086410522461, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8579431772232056, "num_tokens": 511112473.0, "step": 13400 }, { "epoch": 1.7047449433914261, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.922893524169922, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8705617785453796, "num_tokens": 511149839.0, "step": 13401 }, { "epoch": 1.7048721536700167, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.917779922485352, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8595796227455139, "num_tokens": 511183054.0, "step": 13402 }, { "epoch": 1.7049993639486072, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.885431289672852, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8739657402038574, "num_tokens": 511218498.0, "step": 13403 }, { "epoch": 1.7051265742271977, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.84415340423584, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8666039705276489, "num_tokens": 511259776.0, "step": 13404 }, { "epoch": 1.705253784505788, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.910605430603027, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8772724866867065, "num_tokens": 511297331.0, "step": 13405 }, { "epoch": 1.7053809947843785, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.795332908630371, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8827204704284668, "num_tokens": 511331055.0, "step": 13406 }, { "epoch": 1.705508205062969, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.837494850158691, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8883655667304993, "num_tokens": 511368651.0, "step": 13407 }, { "epoch": 1.7056354153415596, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.858491897583008, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8751421570777893, "num_tokens": 511402163.0, "step": 13408 }, { "epoch": 1.7057626256201501, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.863070487976074, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8636506199836731, "num_tokens": 511443769.0, "step": 13409 }, { "epoch": 1.7058898358987407, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.856583595275879, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8676226139068604, "num_tokens": 511481038.0, "step": 13410 }, { "epoch": 1.706017046177331, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.851118087768555, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8717125654220581, "num_tokens": 511522940.0, "step": 13411 }, { "epoch": 1.7061442564559215, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.84838581085205, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8673231601715088, "num_tokens": 511562630.0, "step": 13412 }, { "epoch": 1.706271466734512, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.889730453491211, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8563852906227112, "num_tokens": 511598986.0, "step": 13413 }, { "epoch": 1.7063986770131025, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.884578704833984, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8727560043334961, "num_tokens": 511633582.0, "step": 13414 }, { "epoch": 1.706525887291693, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.876798629760742, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8706888556480408, "num_tokens": 511673296.0, "step": 13415 }, { "epoch": 1.7066530975702836, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.876893043518066, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8617250323295593, "num_tokens": 511707890.0, "step": 13416 }, { "epoch": 1.7067803078488741, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.823321342468262, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8765721321105957, "num_tokens": 511741597.0, "step": 13417 }, { "epoch": 1.7069075181274647, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.873047828674316, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8739330768585205, "num_tokens": 511776816.0, "step": 13418 }, { "epoch": 1.7070347284060552, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.870248794555664, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8646758794784546, "num_tokens": 511814866.0, "step": 13419 }, { "epoch": 1.7071619386846457, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.848383903503418, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8714404702186584, "num_tokens": 511852458.0, "step": 13420 }, { "epoch": 1.7072891489632362, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.913722038269043, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.865128219127655, "num_tokens": 511891201.0, "step": 13421 }, { "epoch": 1.7074163592418268, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.872766494750977, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8703552484512329, "num_tokens": 511930862.0, "step": 13422 }, { "epoch": 1.7075435695204173, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.88309097290039, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8768941164016724, "num_tokens": 511967695.0, "step": 13423 }, { "epoch": 1.7076707797990078, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.874807357788086, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8842256665229797, "num_tokens": 512005976.0, "step": 13424 }, { "epoch": 1.7077979900775984, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.871935844421387, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8717211484909058, "num_tokens": 512041575.0, "step": 13425 }, { "epoch": 1.7079252003561889, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.862504005432129, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8661128282546997, "num_tokens": 512076171.0, "step": 13426 }, { "epoch": 1.7080524106347794, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.910720825195312, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8660427331924438, "num_tokens": 512112082.0, "step": 13427 }, { "epoch": 1.70817962091337, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.810673713684082, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8574585318565369, "num_tokens": 512149126.0, "step": 13428 }, { "epoch": 1.7083068311919605, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.892864227294922, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8708325028419495, "num_tokens": 512182368.0, "step": 13429 }, { "epoch": 1.7084340414705508, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.857535362243652, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8500698208808899, "num_tokens": 512218224.0, "step": 13430 }, { "epoch": 1.7085612517491413, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.951264381408691, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8696656823158264, "num_tokens": 512255523.0, "step": 13431 }, { "epoch": 1.7086884620277318, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.924951553344727, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8798859119415283, "num_tokens": 512290514.0, "step": 13432 }, { "epoch": 1.7088156723063224, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.788963317871094, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8793855905532837, "num_tokens": 512328468.0, "step": 13433 }, { "epoch": 1.708942882584913, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.851865768432617, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8695242404937744, "num_tokens": 512359786.0, "step": 13434 }, { "epoch": 1.7090700928635034, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.83724308013916, "learning_rate": 1e-06, "loss": 0.4891, "mean_token_accuracy": 0.8486630916595459, "num_tokens": 512395679.0, "step": 13435 }, { "epoch": 1.7091973031420937, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.822474479675293, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8688503503799438, "num_tokens": 512433491.0, "step": 13436 }, { "epoch": 1.7093245134206843, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.91746711730957, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.865085780620575, "num_tokens": 512470215.0, "step": 13437 }, { "epoch": 1.7094517236992748, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.832659721374512, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8647060990333557, "num_tokens": 512507055.0, "step": 13438 }, { "epoch": 1.7095789339778653, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.909967422485352, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8518897891044617, "num_tokens": 512547877.0, "step": 13439 }, { "epoch": 1.7097061442564558, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.844588279724121, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.856127142906189, "num_tokens": 512586558.0, "step": 13440 }, { "epoch": 1.7098333545350464, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.809479713439941, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.866851806640625, "num_tokens": 512631181.0, "step": 13441 }, { "epoch": 1.709960564813637, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.7870454788208, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8796547651290894, "num_tokens": 512667925.0, "step": 13442 }, { "epoch": 1.7100877750922274, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.901786804199219, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8513990640640259, "num_tokens": 512703136.0, "step": 13443 }, { "epoch": 1.710214985370818, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.89827823638916, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8725299835205078, "num_tokens": 512739001.0, "step": 13444 }, { "epoch": 1.7103421956494085, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.78526496887207, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8735731840133667, "num_tokens": 512780517.0, "step": 13445 }, { "epoch": 1.710469405927999, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.861310005187988, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8758260011672974, "num_tokens": 512815964.0, "step": 13446 }, { "epoch": 1.7105966162065895, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.8804349899292, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8735399842262268, "num_tokens": 512855451.0, "step": 13447 }, { "epoch": 1.71072382648518, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.865793228149414, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8737056255340576, "num_tokens": 512895809.0, "step": 13448 }, { "epoch": 1.7108510367637706, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.880911827087402, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8745085000991821, "num_tokens": 512936424.0, "step": 13449 }, { "epoch": 1.7109782470423611, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.883641242980957, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8840057849884033, "num_tokens": 512976486.0, "step": 13450 }, { "epoch": 1.7111054573209517, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.800687789916992, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8609322905540466, "num_tokens": 513014131.0, "step": 13451 }, { "epoch": 1.7112326675995422, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.872822761535645, "learning_rate": 1e-06, "loss": 0.4914, "mean_token_accuracy": 0.8421142101287842, "num_tokens": 513054706.0, "step": 13452 }, { "epoch": 1.7113598778781327, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.846251487731934, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8744068145751953, "num_tokens": 513094089.0, "step": 13453 }, { "epoch": 1.711487088156723, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.862014770507812, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8680543303489685, "num_tokens": 513136306.0, "step": 13454 }, { "epoch": 1.7116142984353135, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.910429954528809, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8663364052772522, "num_tokens": 513174376.0, "step": 13455 }, { "epoch": 1.711741508713904, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.884486198425293, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8735366463661194, "num_tokens": 513211834.0, "step": 13456 }, { "epoch": 1.7118687189924946, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.979686737060547, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8620216846466064, "num_tokens": 513241765.0, "step": 13457 }, { "epoch": 1.7119959292710851, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.857449531555176, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8567527532577515, "num_tokens": 513283510.0, "step": 13458 }, { "epoch": 1.7121231395496757, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.976147651672363, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8668127059936523, "num_tokens": 513320746.0, "step": 13459 }, { "epoch": 1.712250349828266, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.917746543884277, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8663032054901123, "num_tokens": 513358036.0, "step": 13460 }, { "epoch": 1.7123775601068565, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.92505931854248, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8750190734863281, "num_tokens": 513395156.0, "step": 13461 }, { "epoch": 1.712504770385447, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.912314414978027, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8630169630050659, "num_tokens": 513438631.0, "step": 13462 }, { "epoch": 1.7126319806640375, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.944622039794922, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8605852723121643, "num_tokens": 513471225.0, "step": 13463 }, { "epoch": 1.712759190942628, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.804903030395508, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8615276217460632, "num_tokens": 513513824.0, "step": 13464 }, { "epoch": 1.7128864012212186, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.916665077209473, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8627694845199585, "num_tokens": 513556263.0, "step": 13465 }, { "epoch": 1.7130136114998091, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.89553165435791, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8731310963630676, "num_tokens": 513595093.0, "step": 13466 }, { "epoch": 1.7131408217783997, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.919700622558594, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8611155152320862, "num_tokens": 513633761.0, "step": 13467 }, { "epoch": 1.7132680320569902, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.929221153259277, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8658287525177002, "num_tokens": 513671404.0, "step": 13468 }, { "epoch": 1.7133952423355807, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.84411334991455, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8678674697875977, "num_tokens": 513713752.0, "step": 13469 }, { "epoch": 1.7135224526141712, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.934126853942871, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8649304509162903, "num_tokens": 513755261.0, "step": 13470 }, { "epoch": 1.7136496628927618, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.908632278442383, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8738415241241455, "num_tokens": 513799042.0, "step": 13471 }, { "epoch": 1.7137768731713523, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.834257125854492, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8591289520263672, "num_tokens": 513838148.0, "step": 13472 }, { "epoch": 1.7139040834499428, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.865281105041504, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.875065267086029, "num_tokens": 513878960.0, "step": 13473 }, { "epoch": 1.7140312937285334, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.914139747619629, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8710508942604065, "num_tokens": 513914083.0, "step": 13474 }, { "epoch": 1.7141585040071239, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.845296859741211, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8576371073722839, "num_tokens": 513951591.0, "step": 13475 }, { "epoch": 1.7142857142857144, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.943838119506836, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.85670006275177, "num_tokens": 513986127.0, "step": 13476 }, { "epoch": 1.714412924564305, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.861194610595703, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8638967871665955, "num_tokens": 514028248.0, "step": 13477 }, { "epoch": 1.7145401348428955, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.933231353759766, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8726018667221069, "num_tokens": 514062757.0, "step": 13478 }, { "epoch": 1.7146673451214858, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.845658302307129, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8556603789329529, "num_tokens": 514102635.0, "step": 13479 }, { "epoch": 1.7147945554000763, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.915643692016602, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8674116134643555, "num_tokens": 514140883.0, "step": 13480 }, { "epoch": 1.7149217656786668, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.945433616638184, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8767938613891602, "num_tokens": 514176536.0, "step": 13481 }, { "epoch": 1.7150489759572574, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.837817192077637, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8632121086120605, "num_tokens": 514209969.0, "step": 13482 }, { "epoch": 1.7151761862358479, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.91657543182373, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8688642978668213, "num_tokens": 514249679.0, "step": 13483 }, { "epoch": 1.7153033965144384, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.942325592041016, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8626465797424316, "num_tokens": 514292719.0, "step": 13484 }, { "epoch": 1.7154306067930287, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.914466857910156, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8641881942749023, "num_tokens": 514323396.0, "step": 13485 }, { "epoch": 1.7155578170716193, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.94314956665039, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8740397691726685, "num_tokens": 514360951.0, "step": 13486 }, { "epoch": 1.7156850273502098, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.9070463180542, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8535914421081543, "num_tokens": 514403879.0, "step": 13487 }, { "epoch": 1.7158122376288003, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.906394958496094, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.875813364982605, "num_tokens": 514446628.0, "step": 13488 }, { "epoch": 1.7159394479073908, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.883744239807129, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8703157901763916, "num_tokens": 514486132.0, "step": 13489 }, { "epoch": 1.7160666581859814, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.87357234954834, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.857823371887207, "num_tokens": 514526384.0, "step": 13490 }, { "epoch": 1.716193868464572, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.87279987335205, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8668823838233948, "num_tokens": 514565828.0, "step": 13491 }, { "epoch": 1.7163210787431624, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 13.032118797302246, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8634935617446899, "num_tokens": 514605260.0, "step": 13492 }, { "epoch": 1.716448289021753, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.946640968322754, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8698466420173645, "num_tokens": 514643433.0, "step": 13493 }, { "epoch": 1.7165754993003435, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.914772987365723, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.870793879032135, "num_tokens": 514683264.0, "step": 13494 }, { "epoch": 1.716702709578934, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.929666519165039, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8751443028450012, "num_tokens": 514719775.0, "step": 13495 }, { "epoch": 1.7168299198575245, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.89907169342041, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8618204593658447, "num_tokens": 514766287.0, "step": 13496 }, { "epoch": 1.716957130136115, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.911023139953613, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8773146271705627, "num_tokens": 514806725.0, "step": 13497 }, { "epoch": 1.7170843404147056, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.851667404174805, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8717629909515381, "num_tokens": 514844008.0, "step": 13498 }, { "epoch": 1.7172115506932961, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.88918685913086, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8577163815498352, "num_tokens": 514880356.0, "step": 13499 }, { "epoch": 1.7173387609718866, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.882671356201172, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8795326948165894, "num_tokens": 514915309.0, "step": 13500 }, { "epoch": 1.7174659712504772, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.943313598632812, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8612068295478821, "num_tokens": 514956496.0, "step": 13501 }, { "epoch": 1.7175931815290677, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.92553424835205, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8690564036369324, "num_tokens": 514992247.0, "step": 13502 }, { "epoch": 1.717720391807658, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.91140365600586, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8603928685188293, "num_tokens": 515035443.0, "step": 13503 }, { "epoch": 1.7178476020862485, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.870539665222168, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8732330799102783, "num_tokens": 515072279.0, "step": 13504 }, { "epoch": 1.717974812364839, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.958632469177246, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8675743937492371, "num_tokens": 515112547.0, "step": 13505 }, { "epoch": 1.7181020226434296, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.882172584533691, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8575085997581482, "num_tokens": 515146814.0, "step": 13506 }, { "epoch": 1.7182292329220201, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.93187427520752, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8710351586341858, "num_tokens": 515184393.0, "step": 13507 }, { "epoch": 1.7183564432006107, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.818985939025879, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8617371320724487, "num_tokens": 515225312.0, "step": 13508 }, { "epoch": 1.718483653479201, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.985735893249512, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8470333218574524, "num_tokens": 515261330.0, "step": 13509 }, { "epoch": 1.7186108637577915, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.880184173583984, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8748396039009094, "num_tokens": 515299325.0, "step": 13510 }, { "epoch": 1.718738074036382, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.898863792419434, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8693351745605469, "num_tokens": 515335213.0, "step": 13511 }, { "epoch": 1.7188652843149725, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.939092636108398, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8802490830421448, "num_tokens": 515365699.0, "step": 13512 }, { "epoch": 1.718992494593563, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.936769485473633, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8781066536903381, "num_tokens": 515396025.0, "step": 13513 }, { "epoch": 1.7191197048721536, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.915992736816406, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8697761297225952, "num_tokens": 515429724.0, "step": 13514 }, { "epoch": 1.7192469151507441, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.949636459350586, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8653777837753296, "num_tokens": 515465894.0, "step": 13515 }, { "epoch": 1.7193741254293347, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.906891822814941, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8872178792953491, "num_tokens": 515498894.0, "step": 13516 }, { "epoch": 1.7195013357079252, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.889281272888184, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8680684566497803, "num_tokens": 515538783.0, "step": 13517 }, { "epoch": 1.7196285459865157, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.935900688171387, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8681832551956177, "num_tokens": 515573107.0, "step": 13518 }, { "epoch": 1.7197557562651062, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.883678436279297, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8703911900520325, "num_tokens": 515614901.0, "step": 13519 }, { "epoch": 1.7198829665436968, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.895440101623535, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8715569972991943, "num_tokens": 515654197.0, "step": 13520 }, { "epoch": 1.7200101768222873, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.87436294555664, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8878539800643921, "num_tokens": 515697003.0, "step": 13521 }, { "epoch": 1.7201373871008778, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.898975372314453, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8656037449836731, "num_tokens": 515737801.0, "step": 13522 }, { "epoch": 1.7202645973794684, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.904250144958496, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8615415096282959, "num_tokens": 515774711.0, "step": 13523 }, { "epoch": 1.7203918076580589, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.891875267028809, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.860967755317688, "num_tokens": 515811728.0, "step": 13524 }, { "epoch": 1.7205190179366494, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.908868789672852, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8750733733177185, "num_tokens": 515844425.0, "step": 13525 }, { "epoch": 1.72064622821524, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.893705368041992, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8710633516311646, "num_tokens": 515882439.0, "step": 13526 }, { "epoch": 1.7207734384938305, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.900197982788086, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8476767539978027, "num_tokens": 515922168.0, "step": 13527 }, { "epoch": 1.7209006487724208, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.963227272033691, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8758750557899475, "num_tokens": 515961840.0, "step": 13528 }, { "epoch": 1.7210278590510113, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.938447952270508, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8691490292549133, "num_tokens": 515997586.0, "step": 13529 }, { "epoch": 1.7211550693296018, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.932321548461914, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8751629590988159, "num_tokens": 516036660.0, "step": 13530 }, { "epoch": 1.7212822796081924, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.951329231262207, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8628376722335815, "num_tokens": 516070846.0, "step": 13531 }, { "epoch": 1.7214094898867829, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.940860748291016, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8896300792694092, "num_tokens": 516107027.0, "step": 13532 }, { "epoch": 1.7215367001653734, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.936307907104492, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8697850108146667, "num_tokens": 516143971.0, "step": 13533 }, { "epoch": 1.7216639104439637, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.90298843383789, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8717919588088989, "num_tokens": 516180617.0, "step": 13534 }, { "epoch": 1.7217911207225542, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.860875129699707, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8697548508644104, "num_tokens": 516220371.0, "step": 13535 }, { "epoch": 1.7219183310011448, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.976758003234863, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8717395663261414, "num_tokens": 516258642.0, "step": 13536 }, { "epoch": 1.7220455412797353, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.777027130126953, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8628406524658203, "num_tokens": 516298971.0, "step": 13537 }, { "epoch": 1.7221727515583258, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.945597648620605, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.880874514579773, "num_tokens": 516334457.0, "step": 13538 }, { "epoch": 1.7222999618369164, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.910162925720215, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8937344551086426, "num_tokens": 516374585.0, "step": 13539 }, { "epoch": 1.7224271721155069, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.941327095031738, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8547060489654541, "num_tokens": 516414180.0, "step": 13540 }, { "epoch": 1.7225543823940974, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.920145988464355, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8779218792915344, "num_tokens": 516450189.0, "step": 13541 }, { "epoch": 1.722681592672688, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.921236038208008, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8672435283660889, "num_tokens": 516487549.0, "step": 13542 }, { "epoch": 1.7228088029512785, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.876044273376465, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8558363914489746, "num_tokens": 516526649.0, "step": 13543 }, { "epoch": 1.722936013229869, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.921902656555176, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8744476437568665, "num_tokens": 516563829.0, "step": 13544 }, { "epoch": 1.7230632235084595, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.001656532287598, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.853486955165863, "num_tokens": 516605179.0, "step": 13545 }, { "epoch": 1.72319043378705, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.91089153289795, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8675101399421692, "num_tokens": 516641576.0, "step": 13546 }, { "epoch": 1.7233176440656406, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.90281867980957, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8669677376747131, "num_tokens": 516678687.0, "step": 13547 }, { "epoch": 1.7234448543442311, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.960905075073242, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8639582395553589, "num_tokens": 516711484.0, "step": 13548 }, { "epoch": 1.7235720646228216, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.907496452331543, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8644132614135742, "num_tokens": 516750598.0, "step": 13549 }, { "epoch": 1.7236992749014122, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.856399536132812, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8756810426712036, "num_tokens": 516793438.0, "step": 13550 }, { "epoch": 1.7238264851800027, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.933290481567383, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8840554356575012, "num_tokens": 516828692.0, "step": 13551 }, { "epoch": 1.723953695458593, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.93594741821289, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8705103397369385, "num_tokens": 516867592.0, "step": 13552 }, { "epoch": 1.7240809057371835, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.92822551727295, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8570302724838257, "num_tokens": 516907668.0, "step": 13553 }, { "epoch": 1.724208116015774, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.849465370178223, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8670419454574585, "num_tokens": 516944628.0, "step": 13554 }, { "epoch": 1.7243353262943646, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.910111427307129, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8641358613967896, "num_tokens": 516989027.0, "step": 13555 }, { "epoch": 1.7244625365729551, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.91248893737793, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8648021221160889, "num_tokens": 517024464.0, "step": 13556 }, { "epoch": 1.7245897468515456, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.873385429382324, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8730483651161194, "num_tokens": 517067714.0, "step": 13557 }, { "epoch": 1.724716957130136, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.918268203735352, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8722773194313049, "num_tokens": 517107564.0, "step": 13558 }, { "epoch": 1.7248441674087265, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.901872634887695, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8744150400161743, "num_tokens": 517148240.0, "step": 13559 }, { "epoch": 1.724971377687317, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.88498306274414, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8616393208503723, "num_tokens": 517187841.0, "step": 13560 }, { "epoch": 1.7250985879659075, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.859940528869629, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8536597490310669, "num_tokens": 517225732.0, "step": 13561 }, { "epoch": 1.725225798244498, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.955107688903809, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8523399829864502, "num_tokens": 517260423.0, "step": 13562 }, { "epoch": 1.7253530085230886, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.930542945861816, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8561331033706665, "num_tokens": 517300862.0, "step": 13563 }, { "epoch": 1.7254802188016791, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.86458969116211, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8734097480773926, "num_tokens": 517341010.0, "step": 13564 }, { "epoch": 1.7256074290802697, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.874663352966309, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8520538806915283, "num_tokens": 517376853.0, "step": 13565 }, { "epoch": 1.7257346393588602, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.880005836486816, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8633196353912354, "num_tokens": 517421854.0, "step": 13566 }, { "epoch": 1.7258618496374507, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.885120391845703, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8624281287193298, "num_tokens": 517461711.0, "step": 13567 }, { "epoch": 1.7259890599160412, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.914578437805176, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8756346702575684, "num_tokens": 517494755.0, "step": 13568 }, { "epoch": 1.7261162701946318, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.897019386291504, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8595188856124878, "num_tokens": 517539061.0, "step": 13569 }, { "epoch": 1.7262434804732223, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.866926193237305, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8636747598648071, "num_tokens": 517579768.0, "step": 13570 }, { "epoch": 1.7263706907518128, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.932231903076172, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8798741698265076, "num_tokens": 517612006.0, "step": 13571 }, { "epoch": 1.7264979010304033, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.957152366638184, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8690106868743896, "num_tokens": 517647547.0, "step": 13572 }, { "epoch": 1.7266251113089939, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.838677406311035, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8594701290130615, "num_tokens": 517689067.0, "step": 13573 }, { "epoch": 1.7267523215875844, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.942891120910645, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8731274008750916, "num_tokens": 517729967.0, "step": 13574 }, { "epoch": 1.726879531866175, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.94006633758545, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8682765960693359, "num_tokens": 517770427.0, "step": 13575 }, { "epoch": 1.7270067421447655, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.817809104919434, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8676389455795288, "num_tokens": 517804202.0, "step": 13576 }, { "epoch": 1.7271339524233558, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.995193481445312, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8609561324119568, "num_tokens": 517834825.0, "step": 13577 }, { "epoch": 1.7272611627019463, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.816766738891602, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8603286147117615, "num_tokens": 517874600.0, "step": 13578 }, { "epoch": 1.7273883729805368, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.91197681427002, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8592382669448853, "num_tokens": 517913488.0, "step": 13579 }, { "epoch": 1.7275155832591274, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.884946823120117, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.881901741027832, "num_tokens": 517951094.0, "step": 13580 }, { "epoch": 1.7276427935377179, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.830387115478516, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8576720356941223, "num_tokens": 517993811.0, "step": 13581 }, { "epoch": 1.7277700038163084, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.91831111907959, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8714392781257629, "num_tokens": 518032860.0, "step": 13582 }, { "epoch": 1.7278972140948987, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.839757919311523, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.864060640335083, "num_tokens": 518070256.0, "step": 13583 }, { "epoch": 1.7280244243734892, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.901897430419922, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8753339648246765, "num_tokens": 518102828.0, "step": 13584 }, { "epoch": 1.7281516346520798, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.890488624572754, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8810765743255615, "num_tokens": 518146859.0, "step": 13585 }, { "epoch": 1.7282788449306703, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.8911771774292, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8750739097595215, "num_tokens": 518186173.0, "step": 13586 }, { "epoch": 1.7284060552092608, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.910219192504883, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8536335825920105, "num_tokens": 518221903.0, "step": 13587 }, { "epoch": 1.7285332654878514, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.905031204223633, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8587661385536194, "num_tokens": 518258048.0, "step": 13588 }, { "epoch": 1.7286604757664419, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.884191513061523, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.859333872795105, "num_tokens": 518299127.0, "step": 13589 }, { "epoch": 1.7287876860450324, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.905145645141602, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.878860354423523, "num_tokens": 518336520.0, "step": 13590 }, { "epoch": 1.728914896323623, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.951190948486328, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8642128109931946, "num_tokens": 518368484.0, "step": 13591 }, { "epoch": 1.7290421066022135, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.902105331420898, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8663516044616699, "num_tokens": 518408439.0, "step": 13592 }, { "epoch": 1.729169316880804, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.8611478805542, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8486319780349731, "num_tokens": 518447414.0, "step": 13593 }, { "epoch": 1.7292965271593945, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.90351676940918, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8652266263961792, "num_tokens": 518486455.0, "step": 13594 }, { "epoch": 1.729423737437985, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.851302146911621, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8849122524261475, "num_tokens": 518527023.0, "step": 13595 }, { "epoch": 1.7295509477165756, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.903566360473633, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8618006706237793, "num_tokens": 518563285.0, "step": 13596 }, { "epoch": 1.729678157995166, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.896971702575684, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8696566820144653, "num_tokens": 518607610.0, "step": 13597 }, { "epoch": 1.7298053682737566, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.925832748413086, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8767192959785461, "num_tokens": 518649434.0, "step": 13598 }, { "epoch": 1.7299325785523472, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.89006519317627, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8748575448989868, "num_tokens": 518684508.0, "step": 13599 }, { "epoch": 1.7300597888309377, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.95062255859375, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8745899200439453, "num_tokens": 518717883.0, "step": 13600 }, { "epoch": 1.730186999109528, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.89268970489502, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8752783536911011, "num_tokens": 518756158.0, "step": 13601 }, { "epoch": 1.7303142093881185, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.965161323547363, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8680130243301392, "num_tokens": 518793003.0, "step": 13602 }, { "epoch": 1.730441419666709, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.953678131103516, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.861316442489624, "num_tokens": 518838485.0, "step": 13603 }, { "epoch": 1.7305686299452996, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.913573265075684, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8623254299163818, "num_tokens": 518885936.0, "step": 13604 }, { "epoch": 1.7306958402238901, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.878463745117188, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8561740517616272, "num_tokens": 518925684.0, "step": 13605 }, { "epoch": 1.7308230505024806, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.926798820495605, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.871147632598877, "num_tokens": 518966438.0, "step": 13606 }, { "epoch": 1.730950260781071, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.892935752868652, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8698428869247437, "num_tokens": 519013576.0, "step": 13607 }, { "epoch": 1.7310774710596615, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.889159202575684, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8656907081604004, "num_tokens": 519045342.0, "step": 13608 }, { "epoch": 1.731204681338252, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.893638610839844, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8540774583816528, "num_tokens": 519088566.0, "step": 13609 }, { "epoch": 1.7313318916168425, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.926334381103516, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8716356158256531, "num_tokens": 519126516.0, "step": 13610 }, { "epoch": 1.731459101895433, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.891648292541504, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8769953846931458, "num_tokens": 519165431.0, "step": 13611 }, { "epoch": 1.7315863121740236, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.920347213745117, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8836336731910706, "num_tokens": 519200991.0, "step": 13612 }, { "epoch": 1.7317135224526141, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.922530174255371, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8745194673538208, "num_tokens": 519245988.0, "step": 13613 }, { "epoch": 1.7318407327312046, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.917692184448242, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8738718032836914, "num_tokens": 519282595.0, "step": 13614 }, { "epoch": 1.7319679430097952, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.922708511352539, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8629741668701172, "num_tokens": 519319061.0, "step": 13615 }, { "epoch": 1.7320951532883857, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.851119995117188, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.866592288017273, "num_tokens": 519358617.0, "step": 13616 }, { "epoch": 1.7322223635669762, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.922072410583496, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8665354251861572, "num_tokens": 519392461.0, "step": 13617 }, { "epoch": 1.7323495738455668, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.899620056152344, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8755276203155518, "num_tokens": 519432893.0, "step": 13618 }, { "epoch": 1.7324767841241573, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.914899826049805, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8683807849884033, "num_tokens": 519473445.0, "step": 13619 }, { "epoch": 1.7326039944027478, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.884688377380371, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8691070079803467, "num_tokens": 519518883.0, "step": 13620 }, { "epoch": 1.7327312046813383, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.842313766479492, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8732890486717224, "num_tokens": 519550123.0, "step": 13621 }, { "epoch": 1.7328584149599289, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.895856857299805, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.859210729598999, "num_tokens": 519592538.0, "step": 13622 }, { "epoch": 1.7329856252385194, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.853540420532227, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8694560527801514, "num_tokens": 519633466.0, "step": 13623 }, { "epoch": 1.73311283551711, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.916802406311035, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8651340007781982, "num_tokens": 519672569.0, "step": 13624 }, { "epoch": 1.7332400457957005, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.880936622619629, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8740427494049072, "num_tokens": 519707383.0, "step": 13625 }, { "epoch": 1.7333672560742908, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.921632766723633, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8499798774719238, "num_tokens": 519748187.0, "step": 13626 }, { "epoch": 1.7334944663528813, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.9585542678833, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8654839992523193, "num_tokens": 519790215.0, "step": 13627 }, { "epoch": 1.7336216766314718, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.916980743408203, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8597434759140015, "num_tokens": 519827278.0, "step": 13628 }, { "epoch": 1.7337488869100623, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.92402172088623, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8744663000106812, "num_tokens": 519860216.0, "step": 13629 }, { "epoch": 1.7338760971886529, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.953564643859863, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8685197830200195, "num_tokens": 519894442.0, "step": 13630 }, { "epoch": 1.7340033074672434, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.944246292114258, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.868301272392273, "num_tokens": 519932524.0, "step": 13631 }, { "epoch": 1.7341305177458337, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.889606475830078, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.867794394493103, "num_tokens": 519974366.0, "step": 13632 }, { "epoch": 1.7342577280244242, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.923852920532227, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8774607181549072, "num_tokens": 520013983.0, "step": 13633 }, { "epoch": 1.7343849383030148, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.965888023376465, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8708889484405518, "num_tokens": 520047671.0, "step": 13634 }, { "epoch": 1.7345121485816053, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.866625785827637, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8643826246261597, "num_tokens": 520086574.0, "step": 13635 }, { "epoch": 1.7346393588601958, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.923105239868164, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8544424176216125, "num_tokens": 520130613.0, "step": 13636 }, { "epoch": 1.7347665691387864, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.874073028564453, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8868818879127502, "num_tokens": 520163763.0, "step": 13637 }, { "epoch": 1.7348937794173769, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.902441024780273, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8652280569076538, "num_tokens": 520198663.0, "step": 13638 }, { "epoch": 1.7350209896959674, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.8985595703125, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8720097541809082, "num_tokens": 520231737.0, "step": 13639 }, { "epoch": 1.735148199974558, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.908458709716797, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8733932971954346, "num_tokens": 520275664.0, "step": 13640 }, { "epoch": 1.7352754102531485, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.924412727355957, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8570663332939148, "num_tokens": 520316100.0, "step": 13641 }, { "epoch": 1.735402620531739, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.999357223510742, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8661485910415649, "num_tokens": 520348167.0, "step": 13642 }, { "epoch": 1.7355298308103295, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.872228622436523, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8550289869308472, "num_tokens": 520385234.0, "step": 13643 }, { "epoch": 1.73565704108892, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.984752655029297, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.850749135017395, "num_tokens": 520423051.0, "step": 13644 }, { "epoch": 1.7357842513675106, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.923576354980469, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8596928119659424, "num_tokens": 520467202.0, "step": 13645 }, { "epoch": 1.735911461646101, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.951560974121094, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8655524253845215, "num_tokens": 520501434.0, "step": 13646 }, { "epoch": 1.7360386719246916, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.958761215209961, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8652288913726807, "num_tokens": 520542757.0, "step": 13647 }, { "epoch": 1.7361658822032822, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.864493370056152, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8597075939178467, "num_tokens": 520585216.0, "step": 13648 }, { "epoch": 1.7362930924818727, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026723861694336, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8690513372421265, "num_tokens": 520620066.0, "step": 13649 }, { "epoch": 1.736420302760463, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.900832176208496, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8747507333755493, "num_tokens": 520660917.0, "step": 13650 }, { "epoch": 1.7365475130390535, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.959015846252441, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8605300784111023, "num_tokens": 520704255.0, "step": 13651 }, { "epoch": 1.736674723317644, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.924117088317871, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8550039529800415, "num_tokens": 520743036.0, "step": 13652 }, { "epoch": 1.7368019335962346, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.97700023651123, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8616605997085571, "num_tokens": 520778325.0, "step": 13653 }, { "epoch": 1.736929143874825, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.92059326171875, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8803290128707886, "num_tokens": 520815226.0, "step": 13654 }, { "epoch": 1.7370563541534156, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.924009323120117, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8487470746040344, "num_tokens": 520856606.0, "step": 13655 }, { "epoch": 1.737183564432006, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.916041374206543, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8596341609954834, "num_tokens": 520893493.0, "step": 13656 }, { "epoch": 1.7373107747105965, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.94322395324707, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.875251293182373, "num_tokens": 520930391.0, "step": 13657 }, { "epoch": 1.737437984989187, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.942934036254883, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8669366836547852, "num_tokens": 520970512.0, "step": 13658 }, { "epoch": 1.7375651952677775, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.980191230773926, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8606066703796387, "num_tokens": 521011787.0, "step": 13659 }, { "epoch": 1.737692405546368, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.992417335510254, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8687397241592407, "num_tokens": 521053494.0, "step": 13660 }, { "epoch": 1.7378196158249586, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.967637062072754, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8568271994590759, "num_tokens": 521089291.0, "step": 13661 }, { "epoch": 1.7379468261035491, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.991137504577637, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8794379234313965, "num_tokens": 521128910.0, "step": 13662 }, { "epoch": 1.7380740363821396, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.902898788452148, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8647692203521729, "num_tokens": 521168175.0, "step": 13663 }, { "epoch": 1.7382012466607302, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.9857816696167, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8631231784820557, "num_tokens": 521208020.0, "step": 13664 }, { "epoch": 1.7383284569393207, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.934943199157715, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8796257376670837, "num_tokens": 521249828.0, "step": 13665 }, { "epoch": 1.7384556672179112, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.990256309509277, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8721452951431274, "num_tokens": 521289940.0, "step": 13666 }, { "epoch": 1.7385828774965018, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.918802261352539, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8852853775024414, "num_tokens": 521327426.0, "step": 13667 }, { "epoch": 1.7387100877750923, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.90335464477539, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8578494787216187, "num_tokens": 521372112.0, "step": 13668 }, { "epoch": 1.7388372980536828, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.964537620544434, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8676177263259888, "num_tokens": 521410842.0, "step": 13669 }, { "epoch": 1.7389645083322733, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.925080299377441, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8778387308120728, "num_tokens": 521446854.0, "step": 13670 }, { "epoch": 1.7390917186108639, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.931577682495117, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8583404421806335, "num_tokens": 521486284.0, "step": 13671 }, { "epoch": 1.7392189288894544, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.986066818237305, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8548411130905151, "num_tokens": 521522716.0, "step": 13672 }, { "epoch": 1.739346139168045, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.986473083496094, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8618872165679932, "num_tokens": 521551549.0, "step": 13673 }, { "epoch": 1.7394733494466355, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.959936141967773, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8707736730575562, "num_tokens": 521589552.0, "step": 13674 }, { "epoch": 1.7396005597252258, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.871841430664062, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8797533512115479, "num_tokens": 521627906.0, "step": 13675 }, { "epoch": 1.7397277700038163, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.888154983520508, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8512596487998962, "num_tokens": 521666615.0, "step": 13676 }, { "epoch": 1.7398549802824068, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.89793586730957, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8630859851837158, "num_tokens": 521707742.0, "step": 13677 }, { "epoch": 1.7399821905609973, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.932435989379883, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8580505847930908, "num_tokens": 521746848.0, "step": 13678 }, { "epoch": 1.7401094008395879, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.906673431396484, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8697762489318848, "num_tokens": 521782917.0, "step": 13679 }, { "epoch": 1.7402366111181784, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.914284706115723, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8647875785827637, "num_tokens": 521825609.0, "step": 13680 }, { "epoch": 1.7403638213967687, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.932172775268555, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8808123469352722, "num_tokens": 521862277.0, "step": 13681 }, { "epoch": 1.7404910316753592, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.92525863647461, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8859935998916626, "num_tokens": 521894951.0, "step": 13682 }, { "epoch": 1.7406182419539498, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.988407135009766, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8673949241638184, "num_tokens": 521932158.0, "step": 13683 }, { "epoch": 1.7407454522325403, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.863438606262207, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8737614750862122, "num_tokens": 521971050.0, "step": 13684 }, { "epoch": 1.7408726625111308, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.948622703552246, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8787603378295898, "num_tokens": 522007617.0, "step": 13685 }, { "epoch": 1.7409998727897213, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.947944641113281, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8563776612281799, "num_tokens": 522053484.0, "step": 13686 }, { "epoch": 1.7411270830683119, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.902052879333496, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8652228713035583, "num_tokens": 522088047.0, "step": 13687 }, { "epoch": 1.7412542933469024, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.95214557647705, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8660391569137573, "num_tokens": 522129865.0, "step": 13688 }, { "epoch": 1.741381503625493, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.926214218139648, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8638544082641602, "num_tokens": 522170154.0, "step": 13689 }, { "epoch": 1.7415087139040835, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.951786994934082, "learning_rate": 1e-06, "loss": 0.344, "mean_token_accuracy": 0.891383945941925, "num_tokens": 522204586.0, "step": 13690 }, { "epoch": 1.741635924182674, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.925934791564941, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.872695803642273, "num_tokens": 522241942.0, "step": 13691 }, { "epoch": 1.7417631344612645, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.878543853759766, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8644426465034485, "num_tokens": 522275800.0, "step": 13692 }, { "epoch": 1.741890344739855, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.961995124816895, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.861871600151062, "num_tokens": 522317024.0, "step": 13693 }, { "epoch": 1.7420175550184456, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.975296974182129, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8777748346328735, "num_tokens": 522358510.0, "step": 13694 }, { "epoch": 1.742144765297036, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.947789192199707, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8617015480995178, "num_tokens": 522395960.0, "step": 13695 }, { "epoch": 1.7422719755756266, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.905896186828613, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8766225576400757, "num_tokens": 522428807.0, "step": 13696 }, { "epoch": 1.7423991858542172, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.967463493347168, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8693129420280457, "num_tokens": 522464472.0, "step": 13697 }, { "epoch": 1.7425263961328077, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.93397331237793, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8814162015914917, "num_tokens": 522509522.0, "step": 13698 }, { "epoch": 1.742653606411398, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.863761901855469, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.852749228477478, "num_tokens": 522545581.0, "step": 13699 }, { "epoch": 1.7427808166899885, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.947368621826172, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8716542720794678, "num_tokens": 522582072.0, "step": 13700 }, { "epoch": 1.742908026968579, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.869296073913574, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8713306188583374, "num_tokens": 522618183.0, "step": 13701 }, { "epoch": 1.7430352372471696, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.916374206542969, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8570928573608398, "num_tokens": 522660578.0, "step": 13702 }, { "epoch": 1.74316244752576, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.949580192565918, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8795060515403748, "num_tokens": 522701934.0, "step": 13703 }, { "epoch": 1.7432896578043506, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.796671867370605, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8858590126037598, "num_tokens": 522737404.0, "step": 13704 }, { "epoch": 1.743416868082941, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.998886108398438, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8648546934127808, "num_tokens": 522774188.0, "step": 13705 }, { "epoch": 1.7435440783615315, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.82275390625, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8770865201950073, "num_tokens": 522810467.0, "step": 13706 }, { "epoch": 1.743671288640122, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.982978820800781, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8572882413864136, "num_tokens": 522845246.0, "step": 13707 }, { "epoch": 1.7437984989187125, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.977672576904297, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.87840735912323, "num_tokens": 522877150.0, "step": 13708 }, { "epoch": 1.743925709197303, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.884564399719238, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8823550939559937, "num_tokens": 522913161.0, "step": 13709 }, { "epoch": 1.7440529194758936, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.913155555725098, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.855208694934845, "num_tokens": 522953464.0, "step": 13710 }, { "epoch": 1.744180129754484, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.882392883300781, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8725533485412598, "num_tokens": 522996874.0, "step": 13711 }, { "epoch": 1.7443073400330746, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.980584144592285, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8722835779190063, "num_tokens": 523031349.0, "step": 13712 }, { "epoch": 1.7444345503116652, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.93243408203125, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8607499599456787, "num_tokens": 523068910.0, "step": 13713 }, { "epoch": 1.7445617605902557, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.937237739562988, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8817192912101746, "num_tokens": 523106179.0, "step": 13714 }, { "epoch": 1.7446889708688462, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.941752433776855, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8707056045532227, "num_tokens": 523140827.0, "step": 13715 }, { "epoch": 1.7448161811474368, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.888690948486328, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8824878334999084, "num_tokens": 523182880.0, "step": 13716 }, { "epoch": 1.7449433914260273, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.988520622253418, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8718889951705933, "num_tokens": 523213059.0, "step": 13717 }, { "epoch": 1.7450706017046178, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.860424995422363, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8677492141723633, "num_tokens": 523254398.0, "step": 13718 }, { "epoch": 1.7451978119832083, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.9315824508667, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8812373280525208, "num_tokens": 523293325.0, "step": 13719 }, { "epoch": 1.7453250222617989, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.899295806884766, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8586264848709106, "num_tokens": 523334402.0, "step": 13720 }, { "epoch": 1.7454522325403894, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.941618919372559, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8729250431060791, "num_tokens": 523376238.0, "step": 13721 }, { "epoch": 1.74557944281898, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.911656379699707, "learning_rate": 1e-06, "loss": 0.3423, "mean_token_accuracy": 0.8913073539733887, "num_tokens": 523417074.0, "step": 13722 }, { "epoch": 1.7457066530975704, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.930339813232422, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8764269351959229, "num_tokens": 523457984.0, "step": 13723 }, { "epoch": 1.7458338633761608, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026754379272461, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8636811375617981, "num_tokens": 523490071.0, "step": 13724 }, { "epoch": 1.7459610736547513, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.885157585144043, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8633734583854675, "num_tokens": 523530105.0, "step": 13725 }, { "epoch": 1.7460882839333418, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.925649642944336, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8821945190429688, "num_tokens": 523576698.0, "step": 13726 }, { "epoch": 1.7462154942119323, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.95167350769043, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8684169054031372, "num_tokens": 523614387.0, "step": 13727 }, { "epoch": 1.7463427044905229, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.963227272033691, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8653333187103271, "num_tokens": 523651149.0, "step": 13728 }, { "epoch": 1.7464699147691134, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.960203170776367, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8811379671096802, "num_tokens": 523688448.0, "step": 13729 }, { "epoch": 1.7465971250477037, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.974617958068848, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8739513158798218, "num_tokens": 523725805.0, "step": 13730 }, { "epoch": 1.7467243353262942, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.974876403808594, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8713207840919495, "num_tokens": 523763564.0, "step": 13731 }, { "epoch": 1.7468515456048848, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.978813171386719, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8640440702438354, "num_tokens": 523804422.0, "step": 13732 }, { "epoch": 1.7469787558834753, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.966483116149902, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8785349726676941, "num_tokens": 523845889.0, "step": 13733 }, { "epoch": 1.7471059661620658, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.962098121643066, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8797287344932556, "num_tokens": 523883623.0, "step": 13734 }, { "epoch": 1.7472331764406563, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.939993858337402, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.875700056552887, "num_tokens": 523917273.0, "step": 13735 }, { "epoch": 1.7473603867192469, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.944902420043945, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8661539554595947, "num_tokens": 523956149.0, "step": 13736 }, { "epoch": 1.7474875969978374, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.947039604187012, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8664100170135498, "num_tokens": 523997229.0, "step": 13737 }, { "epoch": 1.747614807276428, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.88831901550293, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8729738593101501, "num_tokens": 524032921.0, "step": 13738 }, { "epoch": 1.7477420175550185, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.91969108581543, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8710943460464478, "num_tokens": 524065016.0, "step": 13739 }, { "epoch": 1.747869227833609, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.915083885192871, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8578242659568787, "num_tokens": 524107001.0, "step": 13740 }, { "epoch": 1.7479964381121995, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.896631240844727, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8631963133811951, "num_tokens": 524149283.0, "step": 13741 }, { "epoch": 1.74812364839079, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.938562393188477, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8699030876159668, "num_tokens": 524187240.0, "step": 13742 }, { "epoch": 1.7482508586693806, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.969727516174316, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8781901597976685, "num_tokens": 524216793.0, "step": 13743 }, { "epoch": 1.748378068947971, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.950993537902832, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8760650157928467, "num_tokens": 524257629.0, "step": 13744 }, { "epoch": 1.7485052792265616, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.9762601852417, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8803957104682922, "num_tokens": 524291090.0, "step": 13745 }, { "epoch": 1.7486324895051522, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.94172191619873, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8819562196731567, "num_tokens": 524324882.0, "step": 13746 }, { "epoch": 1.7487596997837427, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.942265510559082, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8535220623016357, "num_tokens": 524359422.0, "step": 13747 }, { "epoch": 1.748886910062333, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.98232364654541, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8753575086593628, "num_tokens": 524398383.0, "step": 13748 }, { "epoch": 1.7490141203409235, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.953229904174805, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8657471537590027, "num_tokens": 524437984.0, "step": 13749 }, { "epoch": 1.749141330619514, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.970394134521484, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.875402569770813, "num_tokens": 524471324.0, "step": 13750 }, { "epoch": 1.7492685408981046, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.945094108581543, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8606160879135132, "num_tokens": 524511471.0, "step": 13751 }, { "epoch": 1.749395751176695, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.010193824768066, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8736435770988464, "num_tokens": 524541743.0, "step": 13752 }, { "epoch": 1.7495229614552856, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.967248916625977, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8699806928634644, "num_tokens": 524581470.0, "step": 13753 }, { "epoch": 1.749650171733876, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026997566223145, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8723131418228149, "num_tokens": 524612722.0, "step": 13754 }, { "epoch": 1.7497773820124665, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.983898162841797, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8780364990234375, "num_tokens": 524647640.0, "step": 13755 }, { "epoch": 1.749904592291057, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.970172882080078, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8627560138702393, "num_tokens": 524682366.0, "step": 13756 }, { "epoch": 1.7500318025696475, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.913281440734863, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8697658181190491, "num_tokens": 524724242.0, "step": 13757 }, { "epoch": 1.750159012848238, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.016324043273926, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8554948568344116, "num_tokens": 524757373.0, "step": 13758 }, { "epoch": 1.7502862231268286, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.95384693145752, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8591052293777466, "num_tokens": 524797704.0, "step": 13759 }, { "epoch": 1.750413433405419, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.010830879211426, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8720359802246094, "num_tokens": 524836710.0, "step": 13760 }, { "epoch": 1.7505406436840096, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.019402503967285, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8674527406692505, "num_tokens": 524870224.0, "step": 13761 }, { "epoch": 1.7506678539626002, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.985489845275879, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8720582127571106, "num_tokens": 524905345.0, "step": 13762 }, { "epoch": 1.7507950642411907, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.985628128051758, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8532885313034058, "num_tokens": 524943893.0, "step": 13763 }, { "epoch": 1.7509222745197812, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.908180236816406, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8684725165367126, "num_tokens": 524982924.0, "step": 13764 }, { "epoch": 1.7510494847983717, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.948366165161133, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8676544427871704, "num_tokens": 525016749.0, "step": 13765 }, { "epoch": 1.7511766950769623, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.105122566223145, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8715192675590515, "num_tokens": 525049178.0, "step": 13766 }, { "epoch": 1.7513039053555528, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.044221878051758, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8558019995689392, "num_tokens": 525083456.0, "step": 13767 }, { "epoch": 1.7514311156341433, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.97974681854248, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8710864782333374, "num_tokens": 525120159.0, "step": 13768 }, { "epoch": 1.7515583259127339, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.99742603302002, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.875169038772583, "num_tokens": 525159174.0, "step": 13769 }, { "epoch": 1.7516855361913244, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.982522010803223, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8662464618682861, "num_tokens": 525198373.0, "step": 13770 }, { "epoch": 1.751812746469915, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.033833503723145, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8753392696380615, "num_tokens": 525240618.0, "step": 13771 }, { "epoch": 1.7519399567485054, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.04553508758545, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8836743831634521, "num_tokens": 525274899.0, "step": 13772 }, { "epoch": 1.7520671670270958, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.02637004852295, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.8493671417236328, "num_tokens": 525310144.0, "step": 13773 }, { "epoch": 1.7521943773056863, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.020892143249512, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8790860176086426, "num_tokens": 525348565.0, "step": 13774 }, { "epoch": 1.7523215875842768, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.044448852539062, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8751420378684998, "num_tokens": 525383709.0, "step": 13775 }, { "epoch": 1.7524487978628673, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.007577896118164, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8634883761405945, "num_tokens": 525423983.0, "step": 13776 }, { "epoch": 1.7525760081414579, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.042701721191406, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.875883936882019, "num_tokens": 525457357.0, "step": 13777 }, { "epoch": 1.7527032184200484, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.975123405456543, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.86749267578125, "num_tokens": 525497021.0, "step": 13778 }, { "epoch": 1.7528304286986387, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.988162994384766, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8860384821891785, "num_tokens": 525533592.0, "step": 13779 }, { "epoch": 1.7529576389772292, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.952618598937988, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8613789081573486, "num_tokens": 525572747.0, "step": 13780 }, { "epoch": 1.7530848492558198, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.005655288696289, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8757855296134949, "num_tokens": 525613578.0, "step": 13781 }, { "epoch": 1.7532120595344103, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.974023818969727, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8757679462432861, "num_tokens": 525655047.0, "step": 13782 }, { "epoch": 1.7533392698130008, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.975637435913086, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8665547966957092, "num_tokens": 525689561.0, "step": 13783 }, { "epoch": 1.7534664800915913, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.949183464050293, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8658093214035034, "num_tokens": 525726946.0, "step": 13784 }, { "epoch": 1.7535936903701819, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.996691703796387, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8625026941299438, "num_tokens": 525762089.0, "step": 13785 }, { "epoch": 1.7537209006487724, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.991973876953125, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8755139708518982, "num_tokens": 525796081.0, "step": 13786 }, { "epoch": 1.753848110927363, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.950953483581543, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8588293790817261, "num_tokens": 525829945.0, "step": 13787 }, { "epoch": 1.7539753212059535, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.986834526062012, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8771858215332031, "num_tokens": 525871289.0, "step": 13788 }, { "epoch": 1.754102531484544, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.943049430847168, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8709346055984497, "num_tokens": 525908455.0, "step": 13789 }, { "epoch": 1.7542297417631345, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.977760314941406, "learning_rate": 1e-06, "loss": 0.3404, "mean_token_accuracy": 0.8898608684539795, "num_tokens": 525943734.0, "step": 13790 }, { "epoch": 1.754356952041725, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.857587814331055, "learning_rate": 1e-06, "loss": 0.3389, "mean_token_accuracy": 0.8915743231773376, "num_tokens": 525988070.0, "step": 13791 }, { "epoch": 1.7544841623203156, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.017094612121582, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.884770929813385, "num_tokens": 526027940.0, "step": 13792 }, { "epoch": 1.754611372598906, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.003826141357422, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8794195055961609, "num_tokens": 526061357.0, "step": 13793 }, { "epoch": 1.7547385828774966, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.942151069641113, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.872606635093689, "num_tokens": 526102157.0, "step": 13794 }, { "epoch": 1.7548657931560872, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.983940124511719, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8613269329071045, "num_tokens": 526137241.0, "step": 13795 }, { "epoch": 1.7549930034346777, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.000665664672852, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.862184464931488, "num_tokens": 526169079.0, "step": 13796 }, { "epoch": 1.755120213713268, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.96895980834961, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8743179440498352, "num_tokens": 526212212.0, "step": 13797 }, { "epoch": 1.7552474239918585, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.958999633789062, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8772784471511841, "num_tokens": 526250380.0, "step": 13798 }, { "epoch": 1.755374634270449, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.939879417419434, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.877204418182373, "num_tokens": 526292525.0, "step": 13799 }, { "epoch": 1.7555018445490396, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.972234725952148, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8713585138320923, "num_tokens": 526330427.0, "step": 13800 }, { "epoch": 1.75562905482763, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026345252990723, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8714708089828491, "num_tokens": 526362954.0, "step": 13801 }, { "epoch": 1.7557562651062206, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.974853515625, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8647778034210205, "num_tokens": 526394310.0, "step": 13802 }, { "epoch": 1.755883475384811, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.898005485534668, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.850498378276825, "num_tokens": 526438507.0, "step": 13803 }, { "epoch": 1.7560106856634015, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.008241653442383, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8673544526100159, "num_tokens": 526475124.0, "step": 13804 }, { "epoch": 1.756137895941992, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.884894371032715, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8787127733230591, "num_tokens": 526514936.0, "step": 13805 }, { "epoch": 1.7562651062205825, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.979259490966797, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8639736175537109, "num_tokens": 526548649.0, "step": 13806 }, { "epoch": 1.756392316499173, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.965563774108887, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.85447096824646, "num_tokens": 526589028.0, "step": 13807 }, { "epoch": 1.7565195267777636, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.916754722595215, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8704490661621094, "num_tokens": 526626327.0, "step": 13808 }, { "epoch": 1.756646737056354, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.982736587524414, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8670154809951782, "num_tokens": 526672264.0, "step": 13809 }, { "epoch": 1.7567739473349446, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.96046257019043, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8636488914489746, "num_tokens": 526709979.0, "step": 13810 }, { "epoch": 1.7569011576135352, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.919425964355469, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8821237683296204, "num_tokens": 526745002.0, "step": 13811 }, { "epoch": 1.7570283678921257, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.958712577819824, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8847932815551758, "num_tokens": 526785684.0, "step": 13812 }, { "epoch": 1.7571555781707162, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.999832153320312, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.887358546257019, "num_tokens": 526820246.0, "step": 13813 }, { "epoch": 1.7572827884493067, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.949188232421875, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8745095133781433, "num_tokens": 526854815.0, "step": 13814 }, { "epoch": 1.7574099987278973, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.892204284667969, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8767001628875732, "num_tokens": 526896336.0, "step": 13815 }, { "epoch": 1.7575372090064878, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.949922561645508, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8817000985145569, "num_tokens": 526939586.0, "step": 13816 }, { "epoch": 1.7576644192850783, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.062362670898438, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8752962350845337, "num_tokens": 526975027.0, "step": 13817 }, { "epoch": 1.7577916295636689, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.902142524719238, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.872665286064148, "num_tokens": 527017820.0, "step": 13818 }, { "epoch": 1.7579188398422594, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.945040702819824, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.881827712059021, "num_tokens": 527051647.0, "step": 13819 }, { "epoch": 1.75804605012085, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.9586820602417, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8748409748077393, "num_tokens": 527089590.0, "step": 13820 }, { "epoch": 1.7581732603994404, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.95569133758545, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8672231435775757, "num_tokens": 527126696.0, "step": 13821 }, { "epoch": 1.7583004706780307, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.948973655700684, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8624968528747559, "num_tokens": 527164823.0, "step": 13822 }, { "epoch": 1.7584276809566213, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.963595390319824, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8538033366203308, "num_tokens": 527208032.0, "step": 13823 }, { "epoch": 1.7585548912352118, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.98262882232666, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8716691732406616, "num_tokens": 527242773.0, "step": 13824 }, { "epoch": 1.7586821015138023, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.885910034179688, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8739058375358582, "num_tokens": 527277968.0, "step": 13825 }, { "epoch": 1.7588093117923929, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.02984619140625, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8841474652290344, "num_tokens": 527311077.0, "step": 13826 }, { "epoch": 1.7589365220709834, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.940313339233398, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8764961957931519, "num_tokens": 527340543.0, "step": 13827 }, { "epoch": 1.7590637323495737, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.916664123535156, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8688756227493286, "num_tokens": 527379596.0, "step": 13828 }, { "epoch": 1.7591909426281642, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.927098274230957, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8700085282325745, "num_tokens": 527417873.0, "step": 13829 }, { "epoch": 1.7593181529067548, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.927672386169434, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8723645210266113, "num_tokens": 527458607.0, "step": 13830 }, { "epoch": 1.7594453631853453, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.029476165771484, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8625579476356506, "num_tokens": 527493299.0, "step": 13831 }, { "epoch": 1.7595725734639358, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.928641319274902, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8581915497779846, "num_tokens": 527536065.0, "step": 13832 }, { "epoch": 1.7596997837425263, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.935615539550781, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8556397557258606, "num_tokens": 527576477.0, "step": 13833 }, { "epoch": 1.7598269940211169, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.958108901977539, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8692145347595215, "num_tokens": 527611777.0, "step": 13834 }, { "epoch": 1.7599542042997074, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.037562370300293, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8877428770065308, "num_tokens": 527650236.0, "step": 13835 }, { "epoch": 1.760081414578298, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.966325759887695, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.876583993434906, "num_tokens": 527685486.0, "step": 13836 }, { "epoch": 1.7602086248568884, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.956428527832031, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.857985258102417, "num_tokens": 527725386.0, "step": 13837 }, { "epoch": 1.760335835135479, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.023356437683105, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8579689264297485, "num_tokens": 527756213.0, "step": 13838 }, { "epoch": 1.7604630454140695, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.97665023803711, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8755666017532349, "num_tokens": 527793884.0, "step": 13839 }, { "epoch": 1.76059025569266, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.03265380859375, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.867763876914978, "num_tokens": 527826428.0, "step": 13840 }, { "epoch": 1.7607174659712506, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.968502044677734, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8630855679512024, "num_tokens": 527866691.0, "step": 13841 }, { "epoch": 1.760844676249841, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.976143836975098, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8788899183273315, "num_tokens": 527909482.0, "step": 13842 }, { "epoch": 1.7609718865284316, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.017881393432617, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8585895299911499, "num_tokens": 527947818.0, "step": 13843 }, { "epoch": 1.7610990968070221, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.055773735046387, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8737890124320984, "num_tokens": 527986760.0, "step": 13844 }, { "epoch": 1.7612263070856127, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.96359634399414, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8641420006752014, "num_tokens": 528024742.0, "step": 13845 }, { "epoch": 1.761353517364203, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.065658569335938, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8566184043884277, "num_tokens": 528060781.0, "step": 13846 }, { "epoch": 1.7614807276427935, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.97769546508789, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8721481561660767, "num_tokens": 528099051.0, "step": 13847 }, { "epoch": 1.761607937921384, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.011283874511719, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8641016483306885, "num_tokens": 528138325.0, "step": 13848 }, { "epoch": 1.7617351481999746, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.022107124328613, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8615757822990417, "num_tokens": 528177997.0, "step": 13849 }, { "epoch": 1.761862358478565, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.987590789794922, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8660932779312134, "num_tokens": 528216360.0, "step": 13850 }, { "epoch": 1.7619895687571556, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.0342435836792, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8853135704994202, "num_tokens": 528251100.0, "step": 13851 }, { "epoch": 1.762116779035746, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.961939811706543, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8746283650398254, "num_tokens": 528286298.0, "step": 13852 }, { "epoch": 1.7622439893143365, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.054729461669922, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8858049511909485, "num_tokens": 528322970.0, "step": 13853 }, { "epoch": 1.762371199592927, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.980082511901855, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8863245248794556, "num_tokens": 528355191.0, "step": 13854 }, { "epoch": 1.7624984098715175, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.003475189208984, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8692029118537903, "num_tokens": 528388975.0, "step": 13855 }, { "epoch": 1.762625620150108, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.98232364654541, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8756787776947021, "num_tokens": 528432139.0, "step": 13856 }, { "epoch": 1.7627528304286986, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.037389755249023, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8799858093261719, "num_tokens": 528468315.0, "step": 13857 }, { "epoch": 1.762880040707289, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.953262329101562, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8591864109039307, "num_tokens": 528510368.0, "step": 13858 }, { "epoch": 1.7630072509858796, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.983643531799316, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8731751441955566, "num_tokens": 528545637.0, "step": 13859 }, { "epoch": 1.7631344612644702, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.970174789428711, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8752078413963318, "num_tokens": 528586267.0, "step": 13860 }, { "epoch": 1.7632616715430607, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.896382331848145, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8742265701293945, "num_tokens": 528629640.0, "step": 13861 }, { "epoch": 1.7633888818216512, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.992480278015137, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8563238382339478, "num_tokens": 528665373.0, "step": 13862 }, { "epoch": 1.7635160921002417, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.994811058044434, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8838784694671631, "num_tokens": 528703426.0, "step": 13863 }, { "epoch": 1.7636433023788323, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.986175537109375, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.852565586566925, "num_tokens": 528744886.0, "step": 13864 }, { "epoch": 1.7637705126574228, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.937250137329102, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8828513026237488, "num_tokens": 528779773.0, "step": 13865 }, { "epoch": 1.7638977229360133, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.953338623046875, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.872806966304779, "num_tokens": 528821878.0, "step": 13866 }, { "epoch": 1.7640249332146039, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.999072074890137, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8716864585876465, "num_tokens": 528858056.0, "step": 13867 }, { "epoch": 1.7641521434931944, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.952345848083496, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8670302629470825, "num_tokens": 528903116.0, "step": 13868 }, { "epoch": 1.764279353771785, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.937861442565918, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8643779754638672, "num_tokens": 528941561.0, "step": 13869 }, { "epoch": 1.7644065640503754, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.923840522766113, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8873305320739746, "num_tokens": 528976215.0, "step": 13870 }, { "epoch": 1.7645337743289657, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.899919509887695, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8768441081047058, "num_tokens": 529019280.0, "step": 13871 }, { "epoch": 1.7646609846075563, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.005553245544434, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8742349147796631, "num_tokens": 529060733.0, "step": 13872 }, { "epoch": 1.7647881948861468, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.912311553955078, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8737461566925049, "num_tokens": 529094948.0, "step": 13873 }, { "epoch": 1.7649154051647373, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.925712585449219, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8816731572151184, "num_tokens": 529131374.0, "step": 13874 }, { "epoch": 1.7650426154433279, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.009414672851562, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8678835034370422, "num_tokens": 529169133.0, "step": 13875 }, { "epoch": 1.7651698257219184, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.93826961517334, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8686805963516235, "num_tokens": 529206467.0, "step": 13876 }, { "epoch": 1.7652970360005087, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.94507122039795, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.860944390296936, "num_tokens": 529244891.0, "step": 13877 }, { "epoch": 1.7654242462790992, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.003729820251465, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8721393346786499, "num_tokens": 529277384.0, "step": 13878 }, { "epoch": 1.7655514565576897, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.94304370880127, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8751325011253357, "num_tokens": 529322530.0, "step": 13879 }, { "epoch": 1.7656786668362803, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.949667930603027, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8655577301979065, "num_tokens": 529356345.0, "step": 13880 }, { "epoch": 1.7658058771148708, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.061532974243164, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8722310066223145, "num_tokens": 529391877.0, "step": 13881 }, { "epoch": 1.7659330873934613, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.909586906433105, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8772792220115662, "num_tokens": 529425460.0, "step": 13882 }, { "epoch": 1.7660602976720519, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.95052719116211, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8614251613616943, "num_tokens": 529459660.0, "step": 13883 }, { "epoch": 1.7661875079506424, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.950026512145996, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8625292181968689, "num_tokens": 529501839.0, "step": 13884 }, { "epoch": 1.766314718229233, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.968271255493164, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.868645429611206, "num_tokens": 529540828.0, "step": 13885 }, { "epoch": 1.7664419285078234, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.943133354187012, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8706681728363037, "num_tokens": 529581578.0, "step": 13886 }, { "epoch": 1.766569138786414, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.955523490905762, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8704245686531067, "num_tokens": 529624363.0, "step": 13887 }, { "epoch": 1.7666963490650045, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.910091400146484, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8773946166038513, "num_tokens": 529665286.0, "step": 13888 }, { "epoch": 1.766823559343595, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.98643684387207, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8666847348213196, "num_tokens": 529704150.0, "step": 13889 }, { "epoch": 1.7669507696221856, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.935737609863281, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8731828927993774, "num_tokens": 529742110.0, "step": 13890 }, { "epoch": 1.767077979900776, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.982292175292969, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8701972961425781, "num_tokens": 529776158.0, "step": 13891 }, { "epoch": 1.7672051901793666, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.949831008911133, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.849111795425415, "num_tokens": 529814002.0, "step": 13892 }, { "epoch": 1.7673324004579571, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.95429515838623, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8734869956970215, "num_tokens": 529846696.0, "step": 13893 }, { "epoch": 1.7674596107365477, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.99960708618164, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8672564029693604, "num_tokens": 529881557.0, "step": 13894 }, { "epoch": 1.767586821015138, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.88389778137207, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8581669330596924, "num_tokens": 529924337.0, "step": 13895 }, { "epoch": 1.7677140312937285, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.997939109802246, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8691121935844421, "num_tokens": 529967195.0, "step": 13896 }, { "epoch": 1.767841241572319, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.939005851745605, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8514575958251953, "num_tokens": 530003454.0, "step": 13897 }, { "epoch": 1.7679684518509096, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.988231658935547, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8443419933319092, "num_tokens": 530042577.0, "step": 13898 }, { "epoch": 1.7680956621295, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.973892211914062, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8557970523834229, "num_tokens": 530084903.0, "step": 13899 }, { "epoch": 1.7682228724080906, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.952166557312012, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8659133911132812, "num_tokens": 530117598.0, "step": 13900 }, { "epoch": 1.768350082686681, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.001723289489746, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8729363679885864, "num_tokens": 530151891.0, "step": 13901 }, { "epoch": 1.7684772929652715, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.99139404296875, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8613278865814209, "num_tokens": 530191263.0, "step": 13902 }, { "epoch": 1.768604503243862, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.953524589538574, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8659864664077759, "num_tokens": 530231156.0, "step": 13903 }, { "epoch": 1.7687317135224525, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.966543197631836, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8830444812774658, "num_tokens": 530270089.0, "step": 13904 }, { "epoch": 1.768858923801043, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.953216552734375, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8733917474746704, "num_tokens": 530309808.0, "step": 13905 }, { "epoch": 1.7689861340796336, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.007237434387207, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8587346076965332, "num_tokens": 530357723.0, "step": 13906 }, { "epoch": 1.769113344358224, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.956343650817871, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8568910360336304, "num_tokens": 530389586.0, "step": 13907 }, { "epoch": 1.7692405546368146, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.972965240478516, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8601117134094238, "num_tokens": 530430356.0, "step": 13908 }, { "epoch": 1.7693677649154052, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.969466209411621, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8747102618217468, "num_tokens": 530472859.0, "step": 13909 }, { "epoch": 1.7694949751939957, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.99753475189209, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8575741648674011, "num_tokens": 530515422.0, "step": 13910 }, { "epoch": 1.7696221854725862, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.001023292541504, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8673663139343262, "num_tokens": 530553842.0, "step": 13911 }, { "epoch": 1.7697493957511767, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.998234748840332, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8688547611236572, "num_tokens": 530592593.0, "step": 13912 }, { "epoch": 1.7698766060297673, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.026123046875, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8636430501937866, "num_tokens": 530632309.0, "step": 13913 }, { "epoch": 1.7700038163083578, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.017988204956055, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8652568459510803, "num_tokens": 530667415.0, "step": 13914 }, { "epoch": 1.7701310265869483, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.995512008666992, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8664275407791138, "num_tokens": 530705736.0, "step": 13915 }, { "epoch": 1.7702582368655388, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.998068809509277, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8781861662864685, "num_tokens": 530744931.0, "step": 13916 }, { "epoch": 1.7703854471441294, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.04636001586914, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8809961080551147, "num_tokens": 530784700.0, "step": 13917 }, { "epoch": 1.77051265742272, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.955870628356934, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8858275413513184, "num_tokens": 530821097.0, "step": 13918 }, { "epoch": 1.7706398677013102, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.960382461547852, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8636775016784668, "num_tokens": 530860444.0, "step": 13919 }, { "epoch": 1.7707670779799007, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.006836891174316, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8667406439781189, "num_tokens": 530907304.0, "step": 13920 }, { "epoch": 1.7708942882584913, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.028468132019043, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8616198301315308, "num_tokens": 530945692.0, "step": 13921 }, { "epoch": 1.7710214985370818, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.956514358520508, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8794317245483398, "num_tokens": 530986695.0, "step": 13922 }, { "epoch": 1.7711487088156723, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.010083198547363, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8804111480712891, "num_tokens": 531022656.0, "step": 13923 }, { "epoch": 1.7712759190942629, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.965044021606445, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8783063292503357, "num_tokens": 531056896.0, "step": 13924 }, { "epoch": 1.7714031293728534, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.989995002746582, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.8434249758720398, "num_tokens": 531097286.0, "step": 13925 }, { "epoch": 1.7715303396514437, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.963674545288086, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8775122165679932, "num_tokens": 531136430.0, "step": 13926 }, { "epoch": 1.7716575499300342, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.971351623535156, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8609315156936646, "num_tokens": 531176698.0, "step": 13927 }, { "epoch": 1.7717847602086247, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.98987102508545, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8682247400283813, "num_tokens": 531214343.0, "step": 13928 }, { "epoch": 1.7719119704872153, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.969420433044434, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8813652992248535, "num_tokens": 531252325.0, "step": 13929 }, { "epoch": 1.7720391807658058, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.000391006469727, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8582600355148315, "num_tokens": 531290378.0, "step": 13930 }, { "epoch": 1.7721663910443963, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.940594673156738, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8597118854522705, "num_tokens": 531330406.0, "step": 13931 }, { "epoch": 1.7722936013229869, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.037025451660156, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8698241710662842, "num_tokens": 531367425.0, "step": 13932 }, { "epoch": 1.7724208116015774, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.022059440612793, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8897020220756531, "num_tokens": 531406465.0, "step": 13933 }, { "epoch": 1.772548021880168, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.98316764831543, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8595952987670898, "num_tokens": 531445356.0, "step": 13934 }, { "epoch": 1.7726752321587584, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.028441429138184, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8666919469833374, "num_tokens": 531484737.0, "step": 13935 }, { "epoch": 1.772802442437349, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.973180770874023, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8657828569412231, "num_tokens": 531523453.0, "step": 13936 }, { "epoch": 1.7729296527159395, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026605606079102, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8763640522956848, "num_tokens": 531559627.0, "step": 13937 }, { "epoch": 1.77305686299453, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.027345657348633, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8740023970603943, "num_tokens": 531597677.0, "step": 13938 }, { "epoch": 1.7731840732731206, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.975257873535156, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8772830367088318, "num_tokens": 531642753.0, "step": 13939 }, { "epoch": 1.773311283551711, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.023764610290527, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8620042204856873, "num_tokens": 531688249.0, "step": 13940 }, { "epoch": 1.7734384938303016, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.01013469696045, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8838973641395569, "num_tokens": 531727994.0, "step": 13941 }, { "epoch": 1.7735657041088921, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.985300064086914, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8730639815330505, "num_tokens": 531764428.0, "step": 13942 }, { "epoch": 1.7736929143874827, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.05762767791748, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.86468505859375, "num_tokens": 531806604.0, "step": 13943 }, { "epoch": 1.773820124666073, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.93945026397705, "learning_rate": 1e-06, "loss": 0.3494, "mean_token_accuracy": 0.8872629404067993, "num_tokens": 531845504.0, "step": 13944 }, { "epoch": 1.7739473349446635, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.028839111328125, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8786170482635498, "num_tokens": 531880962.0, "step": 13945 }, { "epoch": 1.774074545223254, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.92823314666748, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8676029443740845, "num_tokens": 531918182.0, "step": 13946 }, { "epoch": 1.7742017555018446, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.054287910461426, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8699713945388794, "num_tokens": 531958340.0, "step": 13947 }, { "epoch": 1.774328965780435, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.019970893859863, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8578172922134399, "num_tokens": 531996951.0, "step": 13948 }, { "epoch": 1.7744561760590256, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.940359115600586, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8733682036399841, "num_tokens": 532036963.0, "step": 13949 }, { "epoch": 1.774583386337616, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0285005569458, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8642674684524536, "num_tokens": 532070937.0, "step": 13950 }, { "epoch": 1.7747105966162064, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.985269546508789, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8560996055603027, "num_tokens": 532107866.0, "step": 13951 }, { "epoch": 1.774837806894797, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.966263771057129, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8620251417160034, "num_tokens": 532151485.0, "step": 13952 }, { "epoch": 1.7749650171733875, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.971168518066406, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8756523728370667, "num_tokens": 532191209.0, "step": 13953 }, { "epoch": 1.775092227451978, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.025010108947754, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8645260334014893, "num_tokens": 532232219.0, "step": 13954 }, { "epoch": 1.7752194377305686, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.0338716506958, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.858153223991394, "num_tokens": 532265330.0, "step": 13955 }, { "epoch": 1.775346648009159, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026352882385254, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8682606816291809, "num_tokens": 532295964.0, "step": 13956 }, { "epoch": 1.7754738582877496, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.994367599487305, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8732935190200806, "num_tokens": 532333804.0, "step": 13957 }, { "epoch": 1.7756010685663401, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.987163543701172, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8733440637588501, "num_tokens": 532373088.0, "step": 13958 }, { "epoch": 1.7757282788449307, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.032464027404785, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8670377135276794, "num_tokens": 532409473.0, "step": 13959 }, { "epoch": 1.7758554891235212, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.044920921325684, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8807097673416138, "num_tokens": 532447883.0, "step": 13960 }, { "epoch": 1.7759826994021117, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.999332427978516, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.873892068862915, "num_tokens": 532485783.0, "step": 13961 }, { "epoch": 1.7761099096807023, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.06898021697998, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8748618364334106, "num_tokens": 532511082.0, "step": 13962 }, { "epoch": 1.7762371199592928, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.036776542663574, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8852618932723999, "num_tokens": 532555516.0, "step": 13963 }, { "epoch": 1.7763643302378833, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.972224235534668, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8542628884315491, "num_tokens": 532599200.0, "step": 13964 }, { "epoch": 1.7764915405164738, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.02199935913086, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8802931308746338, "num_tokens": 532636335.0, "step": 13965 }, { "epoch": 1.7766187507950644, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0361328125, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8810950517654419, "num_tokens": 532678337.0, "step": 13966 }, { "epoch": 1.776745961073655, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.935072898864746, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8655259013175964, "num_tokens": 532715051.0, "step": 13967 }, { "epoch": 1.7768731713522452, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026391983032227, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8801025152206421, "num_tokens": 532750836.0, "step": 13968 }, { "epoch": 1.7770003816308357, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.976592063903809, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8757587671279907, "num_tokens": 532790331.0, "step": 13969 }, { "epoch": 1.7771275919094263, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.992389678955078, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8659587502479553, "num_tokens": 532830169.0, "step": 13970 }, { "epoch": 1.7772548021880168, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.024564743041992, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8622183799743652, "num_tokens": 532865957.0, "step": 13971 }, { "epoch": 1.7773820124666073, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.0482759475708, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8654998540878296, "num_tokens": 532906112.0, "step": 13972 }, { "epoch": 1.7775092227451978, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.984647750854492, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8757027387619019, "num_tokens": 532939676.0, "step": 13973 }, { "epoch": 1.7776364330237884, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.00694465637207, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8518779873847961, "num_tokens": 532977668.0, "step": 13974 }, { "epoch": 1.7777636433023787, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.99522876739502, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8670885562896729, "num_tokens": 533013862.0, "step": 13975 }, { "epoch": 1.7778908535809692, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.039566993713379, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8587256669998169, "num_tokens": 533051222.0, "step": 13976 }, { "epoch": 1.7780180638595597, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.021883010864258, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8675099611282349, "num_tokens": 533090182.0, "step": 13977 }, { "epoch": 1.7781452741381503, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.004557609558105, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8830196857452393, "num_tokens": 533129600.0, "step": 13978 }, { "epoch": 1.7782724844167408, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.018902778625488, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.868762731552124, "num_tokens": 533172351.0, "step": 13979 }, { "epoch": 1.7783996946953313, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.067673683166504, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8668493032455444, "num_tokens": 533211583.0, "step": 13980 }, { "epoch": 1.7785269049739219, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.008240699768066, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8756102323532104, "num_tokens": 533249960.0, "step": 13981 }, { "epoch": 1.7786541152525124, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.037585258483887, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8778403401374817, "num_tokens": 533287328.0, "step": 13982 }, { "epoch": 1.778781325531103, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.966961860656738, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8699076175689697, "num_tokens": 533327772.0, "step": 13983 }, { "epoch": 1.7789085358096934, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.038166999816895, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8627288341522217, "num_tokens": 533372004.0, "step": 13984 }, { "epoch": 1.779035746088284, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.011763572692871, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8681467175483704, "num_tokens": 533409502.0, "step": 13985 }, { "epoch": 1.7791629563668745, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.172245979309082, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8575702905654907, "num_tokens": 533451160.0, "step": 13986 }, { "epoch": 1.779290166645465, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026737213134766, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8772069215774536, "num_tokens": 533480413.0, "step": 13987 }, { "epoch": 1.7794173769240555, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.063920021057129, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8577055335044861, "num_tokens": 533518826.0, "step": 13988 }, { "epoch": 1.779544587202646, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.035979270935059, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8587669730186462, "num_tokens": 533563760.0, "step": 13989 }, { "epoch": 1.7796717974812366, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.108851432800293, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8701469898223877, "num_tokens": 533601499.0, "step": 13990 }, { "epoch": 1.7797990077598271, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.995044708251953, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8730928301811218, "num_tokens": 533640757.0, "step": 13991 }, { "epoch": 1.7799262180384177, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.095815658569336, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8712880611419678, "num_tokens": 533678047.0, "step": 13992 }, { "epoch": 1.780053428317008, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.029791831970215, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8697372674942017, "num_tokens": 533720903.0, "step": 13993 }, { "epoch": 1.7801806385955985, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.092950820922852, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8709734082221985, "num_tokens": 533759720.0, "step": 13994 }, { "epoch": 1.780307848874189, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.052840232849121, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8578687310218811, "num_tokens": 533796122.0, "step": 13995 }, { "epoch": 1.7804350591527796, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.998903274536133, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8752794861793518, "num_tokens": 533836326.0, "step": 13996 }, { "epoch": 1.78056226943137, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.018463134765625, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8802134990692139, "num_tokens": 533867874.0, "step": 13997 }, { "epoch": 1.7806894797099606, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.130526542663574, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8817451000213623, "num_tokens": 533902877.0, "step": 13998 }, { "epoch": 1.780816689988551, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.041927337646484, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8727035522460938, "num_tokens": 533937440.0, "step": 13999 }, { "epoch": 1.7809439002671414, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.99368953704834, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8791913986206055, "num_tokens": 533972308.0, "step": 14000 }, { "epoch": 1.781071110545732, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.997847557067871, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.857292115688324, "num_tokens": 534011913.0, "step": 14001 }, { "epoch": 1.7811983208243225, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.02372932434082, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8807880282402039, "num_tokens": 534057235.0, "step": 14002 }, { "epoch": 1.781325531102913, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.080071449279785, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8692651391029358, "num_tokens": 534092850.0, "step": 14003 }, { "epoch": 1.7814527413815036, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.05865478515625, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8745325803756714, "num_tokens": 534135310.0, "step": 14004 }, { "epoch": 1.781579951660094, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.020808219909668, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8687757849693298, "num_tokens": 534176008.0, "step": 14005 }, { "epoch": 1.7817071619386846, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.063185691833496, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8707516193389893, "num_tokens": 534208217.0, "step": 14006 }, { "epoch": 1.7818343722172751, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.040365219116211, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8742210865020752, "num_tokens": 534247564.0, "step": 14007 }, { "epoch": 1.7819615824958657, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.019925117492676, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8508162498474121, "num_tokens": 534281144.0, "step": 14008 }, { "epoch": 1.7820887927744562, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.094911575317383, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8596961498260498, "num_tokens": 534317893.0, "step": 14009 }, { "epoch": 1.7822160030530467, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.965086936950684, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8675252199172974, "num_tokens": 534356116.0, "step": 14010 }, { "epoch": 1.7823432133316373, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.039549827575684, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.8447757959365845, "num_tokens": 534390499.0, "step": 14011 }, { "epoch": 1.7824704236102278, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.041096687316895, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8644624948501587, "num_tokens": 534424242.0, "step": 14012 }, { "epoch": 1.7825976338888183, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.068099021911621, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.860404372215271, "num_tokens": 534467465.0, "step": 14013 }, { "epoch": 1.7827248441674088, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.973029136657715, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8642971515655518, "num_tokens": 534502959.0, "step": 14014 }, { "epoch": 1.7828520544459994, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.985011100769043, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8698408007621765, "num_tokens": 534549888.0, "step": 14015 }, { "epoch": 1.78297926472459, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.019834518432617, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8719432950019836, "num_tokens": 534591686.0, "step": 14016 }, { "epoch": 1.7831064750031802, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.057394981384277, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8849161863327026, "num_tokens": 534628584.0, "step": 14017 }, { "epoch": 1.7832336852817707, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.112523078918457, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8597773313522339, "num_tokens": 534665684.0, "step": 14018 }, { "epoch": 1.7833608955603613, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.043045043945312, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.871717095375061, "num_tokens": 534701604.0, "step": 14019 }, { "epoch": 1.7834881058389518, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.9954195022583, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8537718653678894, "num_tokens": 534741765.0, "step": 14020 }, { "epoch": 1.7836153161175423, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.095706939697266, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8791791200637817, "num_tokens": 534780331.0, "step": 14021 }, { "epoch": 1.7837425263961328, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.027264595031738, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8578123450279236, "num_tokens": 534816254.0, "step": 14022 }, { "epoch": 1.7838697366747234, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.0322904586792, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8676778674125671, "num_tokens": 534851367.0, "step": 14023 }, { "epoch": 1.7839969469533137, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0137357711792, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8668410778045654, "num_tokens": 534889077.0, "step": 14024 }, { "epoch": 1.7841241572319042, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.0097017288208, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8591997623443604, "num_tokens": 534924249.0, "step": 14025 }, { "epoch": 1.7842513675104947, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.05362606048584, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.854619026184082, "num_tokens": 534966370.0, "step": 14026 }, { "epoch": 1.7843785777890853, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.94459056854248, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8759764432907104, "num_tokens": 535005869.0, "step": 14027 }, { "epoch": 1.7845057880676758, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.986714363098145, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8666988611221313, "num_tokens": 535042057.0, "step": 14028 }, { "epoch": 1.7846329983462663, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.003332138061523, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8633229732513428, "num_tokens": 535074410.0, "step": 14029 }, { "epoch": 1.7847602086248568, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.013517379760742, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8597027063369751, "num_tokens": 535120356.0, "step": 14030 }, { "epoch": 1.7848874189034474, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.007472038269043, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8600119948387146, "num_tokens": 535156939.0, "step": 14031 }, { "epoch": 1.785014629182038, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0350923538208, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8715717792510986, "num_tokens": 535195553.0, "step": 14032 }, { "epoch": 1.7851418394606284, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.07235050201416, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.874245285987854, "num_tokens": 535235692.0, "step": 14033 }, { "epoch": 1.785269049739219, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.014567375183105, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8677210807800293, "num_tokens": 535269265.0, "step": 14034 }, { "epoch": 1.7853962600178095, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.975663185119629, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8843317627906799, "num_tokens": 535303825.0, "step": 14035 }, { "epoch": 1.7855234702964, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.044397354125977, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8646097779273987, "num_tokens": 535338802.0, "step": 14036 }, { "epoch": 1.7856506805749905, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.980354309082031, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8766427040100098, "num_tokens": 535376592.0, "step": 14037 }, { "epoch": 1.785777890853581, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.998104095458984, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8658312559127808, "num_tokens": 535416528.0, "step": 14038 }, { "epoch": 1.7859051011321716, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.007445335388184, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8654865622520447, "num_tokens": 535450892.0, "step": 14039 }, { "epoch": 1.7860323114107621, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.968667984008789, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.871242344379425, "num_tokens": 535488198.0, "step": 14040 }, { "epoch": 1.7861595216893527, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.023722648620605, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8672162890434265, "num_tokens": 535526020.0, "step": 14041 }, { "epoch": 1.786286731967943, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.012518882751465, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8739959001541138, "num_tokens": 535560400.0, "step": 14042 }, { "epoch": 1.7864139422465335, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.031327247619629, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8630161881446838, "num_tokens": 535594699.0, "step": 14043 }, { "epoch": 1.786541152525124, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.004094123840332, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8746587634086609, "num_tokens": 535627019.0, "step": 14044 }, { "epoch": 1.7866683628037145, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.975471496582031, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8791710138320923, "num_tokens": 535667479.0, "step": 14045 }, { "epoch": 1.786795573082305, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.028926849365234, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8630200624465942, "num_tokens": 535705211.0, "step": 14046 }, { "epoch": 1.7869227833608956, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.012063026428223, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8497725129127502, "num_tokens": 535744907.0, "step": 14047 }, { "epoch": 1.787049993639486, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.013147354125977, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8679929971694946, "num_tokens": 535782526.0, "step": 14048 }, { "epoch": 1.7871772039180764, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.974925994873047, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8852795958518982, "num_tokens": 535824153.0, "step": 14049 }, { "epoch": 1.787304414196667, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.018468856811523, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8814959526062012, "num_tokens": 535859019.0, "step": 14050 }, { "epoch": 1.7874316244752575, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.013091087341309, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8809231519699097, "num_tokens": 535896866.0, "step": 14051 }, { "epoch": 1.787558834753848, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.011431694030762, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8703551292419434, "num_tokens": 535937128.0, "step": 14052 }, { "epoch": 1.7876860450324386, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.979918479919434, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8847278952598572, "num_tokens": 535972727.0, "step": 14053 }, { "epoch": 1.787813255311029, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.95976734161377, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8740748167037964, "num_tokens": 536011491.0, "step": 14054 }, { "epoch": 1.7879404655896196, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.987292289733887, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.873229444026947, "num_tokens": 536048074.0, "step": 14055 }, { "epoch": 1.7880676758682101, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.918646812438965, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8758646249771118, "num_tokens": 536089554.0, "step": 14056 }, { "epoch": 1.7881948861468007, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.067376136779785, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8792989253997803, "num_tokens": 536127672.0, "step": 14057 }, { "epoch": 1.7883220964253912, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.982039451599121, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8706552386283875, "num_tokens": 536157517.0, "step": 14058 }, { "epoch": 1.7884493067039817, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.016704559326172, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8824017643928528, "num_tokens": 536192855.0, "step": 14059 }, { "epoch": 1.7885765169825723, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.983358383178711, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8713131546974182, "num_tokens": 536231727.0, "step": 14060 }, { "epoch": 1.7887037272611628, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.000155448913574, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8821980953216553, "num_tokens": 536270738.0, "step": 14061 }, { "epoch": 1.7888309375397533, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.948677062988281, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8876011371612549, "num_tokens": 536313006.0, "step": 14062 }, { "epoch": 1.7889581478183438, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.965229988098145, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.864512026309967, "num_tokens": 536350697.0, "step": 14063 }, { "epoch": 1.7890853580969344, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.00239372253418, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8780423402786255, "num_tokens": 536394252.0, "step": 14064 }, { "epoch": 1.789212568375525, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.045632362365723, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8686085939407349, "num_tokens": 536426591.0, "step": 14065 }, { "epoch": 1.7893397786541152, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.933034896850586, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8584747314453125, "num_tokens": 536468567.0, "step": 14066 }, { "epoch": 1.7894669889327057, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.977300643920898, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8610566854476929, "num_tokens": 536506352.0, "step": 14067 }, { "epoch": 1.7895941992112963, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.053964614868164, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8632948398590088, "num_tokens": 536538122.0, "step": 14068 }, { "epoch": 1.7897214094898868, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.01651382446289, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8600078821182251, "num_tokens": 536574617.0, "step": 14069 }, { "epoch": 1.7898486197684773, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.045928955078125, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8813796043395996, "num_tokens": 536611895.0, "step": 14070 }, { "epoch": 1.7899758300470678, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.985873222351074, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8631656169891357, "num_tokens": 536651336.0, "step": 14071 }, { "epoch": 1.7901030403256584, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.060169219970703, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8660790920257568, "num_tokens": 536692566.0, "step": 14072 }, { "epoch": 1.7902302506042487, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.148375511169434, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.876244068145752, "num_tokens": 536730855.0, "step": 14073 }, { "epoch": 1.7903574608828392, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.005821228027344, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8680840730667114, "num_tokens": 536766478.0, "step": 14074 }, { "epoch": 1.7904846711614297, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.112616539001465, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8644314408302307, "num_tokens": 536799900.0, "step": 14075 }, { "epoch": 1.7906118814400203, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.01229476928711, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8769180774688721, "num_tokens": 536843597.0, "step": 14076 }, { "epoch": 1.7907390917186108, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0263032913208, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8807334899902344, "num_tokens": 536879189.0, "step": 14077 }, { "epoch": 1.7908663019972013, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.038909912109375, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8744361400604248, "num_tokens": 536918099.0, "step": 14078 }, { "epoch": 1.7909935122757918, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0889310836792, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8611444234848022, "num_tokens": 536954410.0, "step": 14079 }, { "epoch": 1.7911207225543824, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.056885719299316, "learning_rate": 1e-06, "loss": 0.3399, "mean_token_accuracy": 0.8920167684555054, "num_tokens": 536986688.0, "step": 14080 }, { "epoch": 1.791247932832973, "ewc_loss": 0.03125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.073967933654785, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8707058429718018, "num_tokens": 537024181.0, "step": 14081 }, { "epoch": 1.7913751431115634, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.99705982208252, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8728518486022949, "num_tokens": 537064174.0, "step": 14082 }, { "epoch": 1.791502353390154, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.055315017700195, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8637840151786804, "num_tokens": 537100279.0, "step": 14083 }, { "epoch": 1.7916295636687445, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.071843147277832, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8459323644638062, "num_tokens": 537135339.0, "step": 14084 }, { "epoch": 1.791756773947335, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.052833557128906, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8647739887237549, "num_tokens": 537169608.0, "step": 14085 }, { "epoch": 1.7918839842259255, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.037346839904785, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8791936635971069, "num_tokens": 537205560.0, "step": 14086 }, { "epoch": 1.792011194504516, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.024638175964355, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8587982058525085, "num_tokens": 537238697.0, "step": 14087 }, { "epoch": 1.7921384047831066, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.05505657196045, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8655685186386108, "num_tokens": 537282365.0, "step": 14088 }, { "epoch": 1.7922656150616971, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.969911575317383, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.879209041595459, "num_tokens": 537320805.0, "step": 14089 }, { "epoch": 1.7923928253402877, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.070306777954102, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8737672567367554, "num_tokens": 537354607.0, "step": 14090 }, { "epoch": 1.792520035618878, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.028473854064941, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8805170059204102, "num_tokens": 537386672.0, "step": 14091 }, { "epoch": 1.7926472458974685, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.000693321228027, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8725535869598389, "num_tokens": 537430882.0, "step": 14092 }, { "epoch": 1.792774456176059, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.017826080322266, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8732446432113647, "num_tokens": 537465114.0, "step": 14093 }, { "epoch": 1.7929016664546495, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.050600051879883, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8644564151763916, "num_tokens": 537502820.0, "step": 14094 }, { "epoch": 1.79302887673324, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.044137954711914, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8794236183166504, "num_tokens": 537537422.0, "step": 14095 }, { "epoch": 1.7931560870118306, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.090943336486816, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8750452995300293, "num_tokens": 537566828.0, "step": 14096 }, { "epoch": 1.793283297290421, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.986014366149902, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8674755096435547, "num_tokens": 537606102.0, "step": 14097 }, { "epoch": 1.7934105075690114, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.1199951171875, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8767111301422119, "num_tokens": 537638527.0, "step": 14098 }, { "epoch": 1.793537717847602, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.978200912475586, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8696048855781555, "num_tokens": 537683375.0, "step": 14099 }, { "epoch": 1.7936649281261925, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.020737648010254, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8651695251464844, "num_tokens": 537731354.0, "step": 14100 }, { "epoch": 1.793792138404783, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.013084411621094, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8654398918151855, "num_tokens": 537770055.0, "step": 14101 }, { "epoch": 1.7939193486833735, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.10658073425293, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.850584864616394, "num_tokens": 537807977.0, "step": 14102 }, { "epoch": 1.794046558961964, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.060477256774902, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8653217554092407, "num_tokens": 537847809.0, "step": 14103 }, { "epoch": 1.7941737692405546, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.00894546508789, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.8504791259765625, "num_tokens": 537885506.0, "step": 14104 }, { "epoch": 1.7943009795191451, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.093642234802246, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8656766414642334, "num_tokens": 537918936.0, "step": 14105 }, { "epoch": 1.7944281897977357, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.01551628112793, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8740071654319763, "num_tokens": 537954205.0, "step": 14106 }, { "epoch": 1.7945554000763262, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.008500099182129, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8718513250350952, "num_tokens": 537989190.0, "step": 14107 }, { "epoch": 1.7946826103549167, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.953913688659668, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8606165051460266, "num_tokens": 538028852.0, "step": 14108 }, { "epoch": 1.7948098206335072, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.030526161193848, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8749387264251709, "num_tokens": 538061245.0, "step": 14109 }, { "epoch": 1.7949370309120978, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.058769226074219, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8820139169692993, "num_tokens": 538103238.0, "step": 14110 }, { "epoch": 1.7950642411906883, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.98326587677002, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8558348417282104, "num_tokens": 538142692.0, "step": 14111 }, { "epoch": 1.7951914514692788, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.086193084716797, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.868047833442688, "num_tokens": 538183155.0, "step": 14112 }, { "epoch": 1.7953186617478694, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.020003318786621, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8685764670372009, "num_tokens": 538222063.0, "step": 14113 }, { "epoch": 1.7954458720264599, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.040999412536621, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8719475269317627, "num_tokens": 538259847.0, "step": 14114 }, { "epoch": 1.7955730823050502, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.979574203491211, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8761451840400696, "num_tokens": 538298557.0, "step": 14115 }, { "epoch": 1.7957002925836407, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.043749809265137, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8671358227729797, "num_tokens": 538337926.0, "step": 14116 }, { "epoch": 1.7958275028622313, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.024258613586426, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.873668909072876, "num_tokens": 538382755.0, "step": 14117 }, { "epoch": 1.7959547131408218, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.992132186889648, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8517835140228271, "num_tokens": 538417639.0, "step": 14118 }, { "epoch": 1.7960819234194123, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.987048149108887, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.864316463470459, "num_tokens": 538458444.0, "step": 14119 }, { "epoch": 1.7962091336980028, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.994988441467285, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8706942796707153, "num_tokens": 538496580.0, "step": 14120 }, { "epoch": 1.7963363439765934, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.009597778320312, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.87892746925354, "num_tokens": 538538474.0, "step": 14121 }, { "epoch": 1.7964635542551837, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.965327262878418, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8719509840011597, "num_tokens": 538576828.0, "step": 14122 }, { "epoch": 1.7965907645337742, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.046348571777344, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8766542673110962, "num_tokens": 538611630.0, "step": 14123 }, { "epoch": 1.7967179748123647, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.983101844787598, "learning_rate": 1e-06, "loss": 0.4949, "mean_token_accuracy": 0.8458963632583618, "num_tokens": 538659537.0, "step": 14124 }, { "epoch": 1.7968451850909553, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0011568069458, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8778302669525146, "num_tokens": 538700216.0, "step": 14125 }, { "epoch": 1.7969723953695458, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.026079177856445, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.867177426815033, "num_tokens": 538737719.0, "step": 14126 }, { "epoch": 1.7970996056481363, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.017082214355469, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.87798672914505, "num_tokens": 538774723.0, "step": 14127 }, { "epoch": 1.7972268159267268, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.045451164245605, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8796489238739014, "num_tokens": 538815288.0, "step": 14128 }, { "epoch": 1.7973540262053174, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.96540355682373, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8685473203659058, "num_tokens": 538858239.0, "step": 14129 }, { "epoch": 1.797481236483908, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.097926139831543, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8727872967720032, "num_tokens": 538896947.0, "step": 14130 }, { "epoch": 1.7976084467624984, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.933767318725586, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.867479681968689, "num_tokens": 538936511.0, "step": 14131 }, { "epoch": 1.797735657041089, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.060981750488281, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8640812635421753, "num_tokens": 538969198.0, "step": 14132 }, { "epoch": 1.7978628673196795, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.054576873779297, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8808532357215881, "num_tokens": 539002258.0, "step": 14133 }, { "epoch": 1.79799007759827, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.03477668762207, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8565313220024109, "num_tokens": 539039993.0, "step": 14134 }, { "epoch": 1.7981172878768605, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.028039932250977, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8660190105438232, "num_tokens": 539080713.0, "step": 14135 }, { "epoch": 1.798244498155451, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.018284797668457, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8800776600837708, "num_tokens": 539121115.0, "step": 14136 }, { "epoch": 1.7983717084340416, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.95923900604248, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.888045608997345, "num_tokens": 539159889.0, "step": 14137 }, { "epoch": 1.7984989187126321, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.961953163146973, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.869168758392334, "num_tokens": 539201508.0, "step": 14138 }, { "epoch": 1.7986261289912227, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.044076919555664, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8520392179489136, "num_tokens": 539238640.0, "step": 14139 }, { "epoch": 1.798753339269813, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.027579307556152, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8659101128578186, "num_tokens": 539274797.0, "step": 14140 }, { "epoch": 1.7988805495484035, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.002520561218262, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.858893632888794, "num_tokens": 539308410.0, "step": 14141 }, { "epoch": 1.799007759826994, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.976134300231934, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8622109889984131, "num_tokens": 539345578.0, "step": 14142 }, { "epoch": 1.7991349701055845, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.079686164855957, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8540516495704651, "num_tokens": 539378946.0, "step": 14143 }, { "epoch": 1.799262180384175, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.918214797973633, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8811829090118408, "num_tokens": 539417663.0, "step": 14144 }, { "epoch": 1.7993893906627656, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.994217872619629, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8841935396194458, "num_tokens": 539459597.0, "step": 14145 }, { "epoch": 1.799516600941356, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.997464179992676, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8715962171554565, "num_tokens": 539499579.0, "step": 14146 }, { "epoch": 1.7996438112199464, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.054275512695312, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8789168000221252, "num_tokens": 539538505.0, "step": 14147 }, { "epoch": 1.799771021498537, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.046041488647461, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8640461564064026, "num_tokens": 539574824.0, "step": 14148 }, { "epoch": 1.7998982317771275, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0663423538208, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8565482497215271, "num_tokens": 539614181.0, "step": 14149 }, { "epoch": 1.800025442055718, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.998819351196289, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.857609748840332, "num_tokens": 539656713.0, "step": 14150 }, { "epoch": 1.8001526523343085, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.065536499023438, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8687309622764587, "num_tokens": 539695436.0, "step": 14151 }, { "epoch": 1.800279862612899, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.997944831848145, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8715735077857971, "num_tokens": 539736549.0, "step": 14152 }, { "epoch": 1.8004070728914896, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.049707412719727, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.875569224357605, "num_tokens": 539776326.0, "step": 14153 }, { "epoch": 1.8005342831700801, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.981508255004883, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8761296272277832, "num_tokens": 539819665.0, "step": 14154 }, { "epoch": 1.8006614934486707, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.038790702819824, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8679083585739136, "num_tokens": 539858175.0, "step": 14155 }, { "epoch": 1.8007887037272612, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.074563980102539, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8688070774078369, "num_tokens": 539895152.0, "step": 14156 }, { "epoch": 1.8009159140058517, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.12692642211914, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8741711378097534, "num_tokens": 539941033.0, "step": 14157 }, { "epoch": 1.8010431242844422, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.982856750488281, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8719133734703064, "num_tokens": 539977959.0, "step": 14158 }, { "epoch": 1.8011703345630328, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.124836921691895, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8680811524391174, "num_tokens": 540014857.0, "step": 14159 }, { "epoch": 1.8012975448416233, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.022395133972168, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8679900765419006, "num_tokens": 540055775.0, "step": 14160 }, { "epoch": 1.8014247551202138, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.071257591247559, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8790426254272461, "num_tokens": 540090287.0, "step": 14161 }, { "epoch": 1.8015519653988044, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.00007152557373, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8775213956832886, "num_tokens": 540129139.0, "step": 14162 }, { "epoch": 1.8016791756773949, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.081807136535645, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8697768449783325, "num_tokens": 540165187.0, "step": 14163 }, { "epoch": 1.8018063859559852, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.989764213562012, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8823277950286865, "num_tokens": 540201113.0, "step": 14164 }, { "epoch": 1.8019335962345757, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.042695999145508, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.871344268321991, "num_tokens": 540237448.0, "step": 14165 }, { "epoch": 1.8020608065131662, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.113452911376953, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8632320761680603, "num_tokens": 540279703.0, "step": 14166 }, { "epoch": 1.8021880167917568, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.028820991516113, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8521850109100342, "num_tokens": 540313071.0, "step": 14167 }, { "epoch": 1.8023152270703473, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.001314163208008, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.873521089553833, "num_tokens": 540353940.0, "step": 14168 }, { "epoch": 1.8024424373489378, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.003719329833984, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8670660257339478, "num_tokens": 540391241.0, "step": 14169 }, { "epoch": 1.8025696476275284, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.058403015136719, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.881624698638916, "num_tokens": 540420205.0, "step": 14170 }, { "epoch": 1.8026968579061187, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.070155143737793, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8732998371124268, "num_tokens": 540452121.0, "step": 14171 }, { "epoch": 1.8028240681847092, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.009040832519531, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8699728846549988, "num_tokens": 540487415.0, "step": 14172 }, { "epoch": 1.8029512784632997, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.087571144104004, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8588038682937622, "num_tokens": 540522293.0, "step": 14173 }, { "epoch": 1.8030784887418903, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.082548141479492, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.882177472114563, "num_tokens": 540560345.0, "step": 14174 }, { "epoch": 1.8032056990204808, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.072758674621582, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8785912990570068, "num_tokens": 540598299.0, "step": 14175 }, { "epoch": 1.8033329092990713, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.017034530639648, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.881146252155304, "num_tokens": 540634480.0, "step": 14176 }, { "epoch": 1.8034601195776618, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.002226829528809, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8609446287155151, "num_tokens": 540679473.0, "step": 14177 }, { "epoch": 1.8035873298562524, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.071462631225586, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8632633686065674, "num_tokens": 540722259.0, "step": 14178 }, { "epoch": 1.803714540134843, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.053112030029297, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8734228610992432, "num_tokens": 540758058.0, "step": 14179 }, { "epoch": 1.8038417504134334, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.050183296203613, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8649373054504395, "num_tokens": 540798535.0, "step": 14180 }, { "epoch": 1.803968960692024, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.051398277282715, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8704502582550049, "num_tokens": 540839423.0, "step": 14181 }, { "epoch": 1.8040961709706145, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.088923454284668, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8763167858123779, "num_tokens": 540887512.0, "step": 14182 }, { "epoch": 1.804223381249205, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.053919792175293, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8720321655273438, "num_tokens": 540922292.0, "step": 14183 }, { "epoch": 1.8043505915277955, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.074344635009766, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8759835362434387, "num_tokens": 540964105.0, "step": 14184 }, { "epoch": 1.804477801806386, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.071981430053711, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8695744276046753, "num_tokens": 541001793.0, "step": 14185 }, { "epoch": 1.8046050120849766, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.03937816619873, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8768860101699829, "num_tokens": 541041104.0, "step": 14186 }, { "epoch": 1.8047322223635671, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.113877296447754, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8636946678161621, "num_tokens": 541079643.0, "step": 14187 }, { "epoch": 1.8048594326421576, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.01413345336914, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8745384812355042, "num_tokens": 541118554.0, "step": 14188 }, { "epoch": 1.804986642920748, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.093713760375977, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8590036630630493, "num_tokens": 541156893.0, "step": 14189 }, { "epoch": 1.8051138531993385, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.036172866821289, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8852062821388245, "num_tokens": 541197283.0, "step": 14190 }, { "epoch": 1.805241063477929, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.001418113708496, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8832006454467773, "num_tokens": 541232343.0, "step": 14191 }, { "epoch": 1.8053682737565195, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.035911560058594, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.861974835395813, "num_tokens": 541270274.0, "step": 14192 }, { "epoch": 1.80549548403511, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.04983139038086, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.864643931388855, "num_tokens": 541306018.0, "step": 14193 }, { "epoch": 1.8056226943137006, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.00467586517334, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.870326817035675, "num_tokens": 541344974.0, "step": 14194 }, { "epoch": 1.805749904592291, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.082335472106934, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8796078562736511, "num_tokens": 541378590.0, "step": 14195 }, { "epoch": 1.8058771148708814, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.077342987060547, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8715973496437073, "num_tokens": 541415945.0, "step": 14196 }, { "epoch": 1.806004325149472, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.970566749572754, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8671359419822693, "num_tokens": 541459577.0, "step": 14197 }, { "epoch": 1.8061315354280625, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.176996231079102, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8703547716140747, "num_tokens": 541501041.0, "step": 14198 }, { "epoch": 1.806258745706653, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.087227821350098, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8638381361961365, "num_tokens": 541538587.0, "step": 14199 }, { "epoch": 1.8063859559852435, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.018779754638672, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8572625517845154, "num_tokens": 541582225.0, "step": 14200 }, { "epoch": 1.806513166263834, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.148177146911621, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8799167275428772, "num_tokens": 541616956.0, "step": 14201 }, { "epoch": 1.8066403765424246, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.107114791870117, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8583984971046448, "num_tokens": 541650814.0, "step": 14202 }, { "epoch": 1.8067675868210151, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.031579971313477, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8794443607330322, "num_tokens": 541692167.0, "step": 14203 }, { "epoch": 1.8068947970996057, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.03943920135498, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8608628511428833, "num_tokens": 541732427.0, "step": 14204 }, { "epoch": 1.8070220073781962, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.020466804504395, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8712319135665894, "num_tokens": 541773256.0, "step": 14205 }, { "epoch": 1.8071492176567867, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 13.073126792907715, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8759373426437378, "num_tokens": 541815008.0, "step": 14206 }, { "epoch": 1.8072764279353772, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.12966537475586, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8729968070983887, "num_tokens": 541853462.0, "step": 14207 }, { "epoch": 1.8074036382139678, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.064241409301758, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8652583360671997, "num_tokens": 541890487.0, "step": 14208 }, { "epoch": 1.8075308484925583, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.032639503479004, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8745666146278381, "num_tokens": 541929857.0, "step": 14209 }, { "epoch": 1.8076580587711488, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.09199333190918, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8642098903656006, "num_tokens": 541969923.0, "step": 14210 }, { "epoch": 1.8077852690497394, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.00139045715332, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8785582184791565, "num_tokens": 542008802.0, "step": 14211 }, { "epoch": 1.8079124793283299, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.029343605041504, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.883148729801178, "num_tokens": 542049276.0, "step": 14212 }, { "epoch": 1.8080396896069202, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.065381050109863, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.863608717918396, "num_tokens": 542086921.0, "step": 14213 }, { "epoch": 1.8081668998855107, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.027953147888184, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8704635500907898, "num_tokens": 542124291.0, "step": 14214 }, { "epoch": 1.8082941101641012, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.043919563293457, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8592343330383301, "num_tokens": 542167600.0, "step": 14215 }, { "epoch": 1.8084213204426918, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.166133880615234, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8660922050476074, "num_tokens": 542201134.0, "step": 14216 }, { "epoch": 1.8085485307212823, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.081008911132812, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8569052219390869, "num_tokens": 542242590.0, "step": 14217 }, { "epoch": 1.8086757409998728, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.059154510498047, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8697916269302368, "num_tokens": 542281945.0, "step": 14218 }, { "epoch": 1.8088029512784631, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.074139595031738, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8781455755233765, "num_tokens": 542319389.0, "step": 14219 }, { "epoch": 1.8089301615570537, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.067627906799316, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8577142953872681, "num_tokens": 542356912.0, "step": 14220 }, { "epoch": 1.8090573718356442, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.070231437683105, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8637360334396362, "num_tokens": 542394054.0, "step": 14221 }, { "epoch": 1.8091845821142347, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.04278564453125, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8753219842910767, "num_tokens": 542431063.0, "step": 14222 }, { "epoch": 1.8093117923928252, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.070250511169434, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8586154580116272, "num_tokens": 542471398.0, "step": 14223 }, { "epoch": 1.8094390026714158, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.015275955200195, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8555858731269836, "num_tokens": 542516288.0, "step": 14224 }, { "epoch": 1.8095662129500063, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.113984107971191, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8683304786682129, "num_tokens": 542552292.0, "step": 14225 }, { "epoch": 1.8096934232285968, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.015835762023926, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8720110058784485, "num_tokens": 542592380.0, "step": 14226 }, { "epoch": 1.8098206335071874, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.068402290344238, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8628183603286743, "num_tokens": 542632698.0, "step": 14227 }, { "epoch": 1.8099478437857779, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.068676948547363, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8583669662475586, "num_tokens": 542668234.0, "step": 14228 }, { "epoch": 1.8100750540643684, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.996623992919922, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8685483932495117, "num_tokens": 542706205.0, "step": 14229 }, { "epoch": 1.810202264342959, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.0547513961792, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8678165078163147, "num_tokens": 542737068.0, "step": 14230 }, { "epoch": 1.8103294746215495, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.020474433898926, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8809664249420166, "num_tokens": 542769776.0, "step": 14231 }, { "epoch": 1.81045668490014, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.030980110168457, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8825087547302246, "num_tokens": 542802440.0, "step": 14232 }, { "epoch": 1.8105838951787305, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.05048942565918, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8813073039054871, "num_tokens": 542841158.0, "step": 14233 }, { "epoch": 1.810711105457321, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026867866516113, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8656017780303955, "num_tokens": 542876637.0, "step": 14234 }, { "epoch": 1.8108383157359116, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.0679349899292, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8763458728790283, "num_tokens": 542915087.0, "step": 14235 }, { "epoch": 1.8109655260145021, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.077461242675781, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8610419034957886, "num_tokens": 542950635.0, "step": 14236 }, { "epoch": 1.8110927362930926, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.11906909942627, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8650202751159668, "num_tokens": 542990078.0, "step": 14237 }, { "epoch": 1.811219946571683, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.094735145568848, "learning_rate": 1e-06, "loss": 0.48, "mean_token_accuracy": 0.8451308012008667, "num_tokens": 543030333.0, "step": 14238 }, { "epoch": 1.8113471568502735, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.024189949035645, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8626279830932617, "num_tokens": 543070537.0, "step": 14239 }, { "epoch": 1.811474367128864, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.056770324707031, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8680340051651001, "num_tokens": 543109137.0, "step": 14240 }, { "epoch": 1.8116015774074545, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.088144302368164, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8632322549819946, "num_tokens": 543146616.0, "step": 14241 }, { "epoch": 1.811728787686045, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.026708602905273, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.882330060005188, "num_tokens": 543178433.0, "step": 14242 }, { "epoch": 1.8118559979646356, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.027021408081055, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8817542195320129, "num_tokens": 543215086.0, "step": 14243 }, { "epoch": 1.811983208243226, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.101295471191406, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8537052869796753, "num_tokens": 543252111.0, "step": 14244 }, { "epoch": 1.8121104185218164, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.026050567626953, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.87724369764328, "num_tokens": 543291938.0, "step": 14245 }, { "epoch": 1.812237628800407, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.00776481628418, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8742013573646545, "num_tokens": 543335141.0, "step": 14246 }, { "epoch": 1.8123648390789975, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.069095611572266, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8700687289237976, "num_tokens": 543371305.0, "step": 14247 }, { "epoch": 1.812492049357588, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.029393196105957, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8804100751876831, "num_tokens": 543407431.0, "step": 14248 }, { "epoch": 1.8126192596361785, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.146120071411133, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8701145648956299, "num_tokens": 543450287.0, "step": 14249 }, { "epoch": 1.812746469914769, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.01286792755127, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8826885223388672, "num_tokens": 543486521.0, "step": 14250 }, { "epoch": 1.8128736801933596, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0802001953125, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8688029646873474, "num_tokens": 543522157.0, "step": 14251 }, { "epoch": 1.8130008904719501, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.124224662780762, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8727315664291382, "num_tokens": 543560503.0, "step": 14252 }, { "epoch": 1.8131281007505406, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.002124786376953, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8582082986831665, "num_tokens": 543598943.0, "step": 14253 }, { "epoch": 1.8132553110291312, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.012747764587402, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8718701601028442, "num_tokens": 543640466.0, "step": 14254 }, { "epoch": 1.8133825213077217, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.010565757751465, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8711683750152588, "num_tokens": 543677581.0, "step": 14255 }, { "epoch": 1.8135097315863122, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.090127944946289, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8625284433364868, "num_tokens": 543711220.0, "step": 14256 }, { "epoch": 1.8136369418649028, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.096216201782227, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8631894588470459, "num_tokens": 543750109.0, "step": 14257 }, { "epoch": 1.8137641521434933, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.042031288146973, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8738243579864502, "num_tokens": 543788040.0, "step": 14258 }, { "epoch": 1.8138913624220838, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0298433303833, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8689873218536377, "num_tokens": 543821037.0, "step": 14259 }, { "epoch": 1.8140185727006743, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.051390647888184, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8829219341278076, "num_tokens": 543856331.0, "step": 14260 }, { "epoch": 1.8141457829792649, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.02637767791748, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8743942975997925, "num_tokens": 543892898.0, "step": 14261 }, { "epoch": 1.8142729932578552, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.061298370361328, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8653368353843689, "num_tokens": 543932858.0, "step": 14262 }, { "epoch": 1.8144002035364457, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.00820541381836, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8790867328643799, "num_tokens": 543968145.0, "step": 14263 }, { "epoch": 1.8145274138150362, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.028387069702148, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.868843674659729, "num_tokens": 544000495.0, "step": 14264 }, { "epoch": 1.8146546240936268, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.073762893676758, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8648593425750732, "num_tokens": 544041345.0, "step": 14265 }, { "epoch": 1.8147818343722173, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.062580108642578, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8745763301849365, "num_tokens": 544080084.0, "step": 14266 }, { "epoch": 1.8149090446508078, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.045599937438965, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8622134923934937, "num_tokens": 544115828.0, "step": 14267 }, { "epoch": 1.8150362549293981, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.033885955810547, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8672447204589844, "num_tokens": 544153213.0, "step": 14268 }, { "epoch": 1.8151634652079887, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.059761047363281, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8729915618896484, "num_tokens": 544192383.0, "step": 14269 }, { "epoch": 1.8152906754865792, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.959817886352539, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.871805727481842, "num_tokens": 544232034.0, "step": 14270 }, { "epoch": 1.8154178857651697, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.078240394592285, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8545013666152954, "num_tokens": 544266242.0, "step": 14271 }, { "epoch": 1.8155450960437602, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.027177810668945, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8595924377441406, "num_tokens": 544304995.0, "step": 14272 }, { "epoch": 1.8156723063223508, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.112278938293457, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.868577241897583, "num_tokens": 544341950.0, "step": 14273 }, { "epoch": 1.8157995166009413, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.062621116638184, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8656218647956848, "num_tokens": 544379535.0, "step": 14274 }, { "epoch": 1.8159267268795318, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.984732627868652, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8687446117401123, "num_tokens": 544416939.0, "step": 14275 }, { "epoch": 1.8160539371581224, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.063542366027832, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8807171583175659, "num_tokens": 544461476.0, "step": 14276 }, { "epoch": 1.8161811474367129, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.053447723388672, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8702221512794495, "num_tokens": 544505051.0, "step": 14277 }, { "epoch": 1.8163083577153034, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.058204650878906, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8696227073669434, "num_tokens": 544550408.0, "step": 14278 }, { "epoch": 1.816435567993894, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.049275398254395, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8689984679222107, "num_tokens": 544586645.0, "step": 14279 }, { "epoch": 1.8165627782724845, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.010843276977539, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8633042573928833, "num_tokens": 544628087.0, "step": 14280 }, { "epoch": 1.816689988551075, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.091612815856934, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8842511773109436, "num_tokens": 544667961.0, "step": 14281 }, { "epoch": 1.8168171988296655, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.059345245361328, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8672236800193787, "num_tokens": 544704919.0, "step": 14282 }, { "epoch": 1.816944409108256, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.086286544799805, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8740612268447876, "num_tokens": 544741421.0, "step": 14283 }, { "epoch": 1.8170716193868466, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.105620384216309, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.873969554901123, "num_tokens": 544776414.0, "step": 14284 }, { "epoch": 1.817198829665437, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.025492668151855, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8630650043487549, "num_tokens": 544814964.0, "step": 14285 }, { "epoch": 1.8173260399440276, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.126134872436523, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8558862805366516, "num_tokens": 544853381.0, "step": 14286 }, { "epoch": 1.817453250222618, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.000921249389648, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8727983236312866, "num_tokens": 544892474.0, "step": 14287 }, { "epoch": 1.8175804605012085, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.09760856628418, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8768643140792847, "num_tokens": 544927667.0, "step": 14288 }, { "epoch": 1.817707670779799, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.129405975341797, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.87549889087677, "num_tokens": 544959780.0, "step": 14289 }, { "epoch": 1.8178348810583895, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.999812126159668, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8630358576774597, "num_tokens": 545001832.0, "step": 14290 }, { "epoch": 1.81796209133698, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.063436508178711, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8779151439666748, "num_tokens": 545038489.0, "step": 14291 }, { "epoch": 1.8180893016155706, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.067481994628906, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.8496789932250977, "num_tokens": 545069086.0, "step": 14292 }, { "epoch": 1.818216511894161, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.068402290344238, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8730471134185791, "num_tokens": 545104564.0, "step": 14293 }, { "epoch": 1.8183437221727514, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.047374725341797, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8731420040130615, "num_tokens": 545149036.0, "step": 14294 }, { "epoch": 1.818470932451342, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.060147285461426, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8814069032669067, "num_tokens": 545183825.0, "step": 14295 }, { "epoch": 1.8185981427299325, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.016186714172363, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8737629652023315, "num_tokens": 545221274.0, "step": 14296 }, { "epoch": 1.818725353008523, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.038545608520508, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8770171403884888, "num_tokens": 545259242.0, "step": 14297 }, { "epoch": 1.8188525632871135, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.117382049560547, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.856277346611023, "num_tokens": 545289380.0, "step": 14298 }, { "epoch": 1.818979773565704, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.099302291870117, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8729776740074158, "num_tokens": 545321935.0, "step": 14299 }, { "epoch": 1.8191069838442946, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.976852416992188, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8773203492164612, "num_tokens": 545365714.0, "step": 14300 }, { "epoch": 1.8192341941228851, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.063699722290039, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8600406050682068, "num_tokens": 545406838.0, "step": 14301 }, { "epoch": 1.8193614044014756, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.06586742401123, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8672515153884888, "num_tokens": 545443265.0, "step": 14302 }, { "epoch": 1.8194886146800662, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.013699531555176, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.880636990070343, "num_tokens": 545479903.0, "step": 14303 }, { "epoch": 1.8196158249586567, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.07939338684082, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8767542243003845, "num_tokens": 545513767.0, "step": 14304 }, { "epoch": 1.8197430352372472, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.043757438659668, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8551756143569946, "num_tokens": 545550541.0, "step": 14305 }, { "epoch": 1.8198702455158378, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.064911842346191, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8709308505058289, "num_tokens": 545590739.0, "step": 14306 }, { "epoch": 1.8199974557944283, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.047924995422363, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8566378355026245, "num_tokens": 545627127.0, "step": 14307 }, { "epoch": 1.8201246660730188, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.059225082397461, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8739044666290283, "num_tokens": 545665473.0, "step": 14308 }, { "epoch": 1.8202518763516093, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.050740242004395, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8851593136787415, "num_tokens": 545708018.0, "step": 14309 }, { "epoch": 1.8203790866301999, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.125725746154785, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.875289261341095, "num_tokens": 545745635.0, "step": 14310 }, { "epoch": 1.8205062969087902, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.019412994384766, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8590680360794067, "num_tokens": 545785260.0, "step": 14311 }, { "epoch": 1.8206335071873807, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.108059883117676, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.868047833442688, "num_tokens": 545821696.0, "step": 14312 }, { "epoch": 1.8207607174659712, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.011006355285645, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8835116624832153, "num_tokens": 545857971.0, "step": 14313 }, { "epoch": 1.8208879277445618, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.159285545349121, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8597730398178101, "num_tokens": 545891000.0, "step": 14314 }, { "epoch": 1.8210151380231523, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.971891403198242, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8567138314247131, "num_tokens": 545932825.0, "step": 14315 }, { "epoch": 1.8211423483017428, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.129436492919922, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8602109551429749, "num_tokens": 545972849.0, "step": 14316 }, { "epoch": 1.8212695585803331, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.09338665008545, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8826576471328735, "num_tokens": 546011492.0, "step": 14317 }, { "epoch": 1.8213967688589237, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.08210277557373, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8650637865066528, "num_tokens": 546048461.0, "step": 14318 }, { "epoch": 1.8215239791375142, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0191068649292, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8792585134506226, "num_tokens": 546087113.0, "step": 14319 }, { "epoch": 1.8216511894161047, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.985089302062988, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8731972575187683, "num_tokens": 546124413.0, "step": 14320 }, { "epoch": 1.8217783996946952, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.067215919494629, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.880804717540741, "num_tokens": 546165763.0, "step": 14321 }, { "epoch": 1.8219056099732858, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.95085620880127, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8705990314483643, "num_tokens": 546203218.0, "step": 14322 }, { "epoch": 1.8220328202518763, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.080124855041504, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8603885769844055, "num_tokens": 546247610.0, "step": 14323 }, { "epoch": 1.8221600305304668, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.002938270568848, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8617614507675171, "num_tokens": 546288952.0, "step": 14324 }, { "epoch": 1.8222872408090574, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.081988334655762, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8808894157409668, "num_tokens": 546322449.0, "step": 14325 }, { "epoch": 1.8224144510876479, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.02984619140625, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8764188289642334, "num_tokens": 546362692.0, "step": 14326 }, { "epoch": 1.8225416613662384, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.052791595458984, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8542313575744629, "num_tokens": 546398856.0, "step": 14327 }, { "epoch": 1.822668871644829, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.999147415161133, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8619188666343689, "num_tokens": 546439633.0, "step": 14328 }, { "epoch": 1.8227960819234195, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.019436836242676, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8737829923629761, "num_tokens": 546477282.0, "step": 14329 }, { "epoch": 1.82292329220201, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.0258207321167, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8811050653457642, "num_tokens": 546514584.0, "step": 14330 }, { "epoch": 1.8230505024806005, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.036575317382812, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8740960955619812, "num_tokens": 546550761.0, "step": 14331 }, { "epoch": 1.823177712759191, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.982335090637207, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8566257953643799, "num_tokens": 546594157.0, "step": 14332 }, { "epoch": 1.8233049230377816, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.057611465454102, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8777080774307251, "num_tokens": 546634671.0, "step": 14333 }, { "epoch": 1.823432133316372, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.030313491821289, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8699228167533875, "num_tokens": 546672084.0, "step": 14334 }, { "epoch": 1.8235593435949626, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.048600196838379, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8643312454223633, "num_tokens": 546701920.0, "step": 14335 }, { "epoch": 1.823686553873553, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.123294830322266, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.874786376953125, "num_tokens": 546735776.0, "step": 14336 }, { "epoch": 1.8238137641521435, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.993045806884766, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.87505042552948, "num_tokens": 546772991.0, "step": 14337 }, { "epoch": 1.823940974430734, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.075586318969727, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8641145825386047, "num_tokens": 546805310.0, "step": 14338 }, { "epoch": 1.8240681847093245, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.098075866699219, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.876110315322876, "num_tokens": 546838240.0, "step": 14339 }, { "epoch": 1.824195394987915, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.048929214477539, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8650895357131958, "num_tokens": 546878484.0, "step": 14340 }, { "epoch": 1.8243226052665056, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.058042526245117, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8850369453430176, "num_tokens": 546919747.0, "step": 14341 }, { "epoch": 1.8244498155450959, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.000422477722168, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8695663213729858, "num_tokens": 546962614.0, "step": 14342 }, { "epoch": 1.8245770258236864, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.031861305236816, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.863821268081665, "num_tokens": 547000059.0, "step": 14343 }, { "epoch": 1.824704236102277, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.997933387756348, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8714753985404968, "num_tokens": 547037372.0, "step": 14344 }, { "epoch": 1.8248314463808675, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.108917236328125, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8566203713417053, "num_tokens": 547071840.0, "step": 14345 }, { "epoch": 1.824958656659458, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.032567024230957, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8463854789733887, "num_tokens": 547108916.0, "step": 14346 }, { "epoch": 1.8250858669380485, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.070967674255371, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8880971670150757, "num_tokens": 547146991.0, "step": 14347 }, { "epoch": 1.825213077216639, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.056958198547363, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8801053762435913, "num_tokens": 547184093.0, "step": 14348 }, { "epoch": 1.8253402874952296, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.041081428527832, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8759894967079163, "num_tokens": 547217570.0, "step": 14349 }, { "epoch": 1.8254674977738201, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.990615844726562, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8695628046989441, "num_tokens": 547259498.0, "step": 14350 }, { "epoch": 1.8255947080524106, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.057917594909668, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8684476017951965, "num_tokens": 547298419.0, "step": 14351 }, { "epoch": 1.8257219183310012, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.981151580810547, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8776804208755493, "num_tokens": 547342984.0, "step": 14352 }, { "epoch": 1.8258491286095917, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.082663536071777, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.87241530418396, "num_tokens": 547383162.0, "step": 14353 }, { "epoch": 1.8259763388881822, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.07433032989502, "learning_rate": 1e-06, "loss": 0.3409, "mean_token_accuracy": 0.8903442025184631, "num_tokens": 547416967.0, "step": 14354 }, { "epoch": 1.8261035491667728, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.011383056640625, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8734245896339417, "num_tokens": 547454695.0, "step": 14355 }, { "epoch": 1.8262307594453633, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.061798095703125, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8801087737083435, "num_tokens": 547494258.0, "step": 14356 }, { "epoch": 1.8263579697239538, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.06323528289795, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8654938340187073, "num_tokens": 547531404.0, "step": 14357 }, { "epoch": 1.8264851800025443, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.032344818115234, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.874836802482605, "num_tokens": 547566553.0, "step": 14358 }, { "epoch": 1.8266123902811349, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.082088470458984, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8634090423583984, "num_tokens": 547605505.0, "step": 14359 }, { "epoch": 1.8267396005597252, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.08991527557373, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8784816265106201, "num_tokens": 547643146.0, "step": 14360 }, { "epoch": 1.8268668108383157, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.003125190734863, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8673677444458008, "num_tokens": 547683373.0, "step": 14361 }, { "epoch": 1.8269940211169062, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.079874992370605, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8654447197914124, "num_tokens": 547716762.0, "step": 14362 }, { "epoch": 1.8271212313954968, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.080865859985352, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8784825205802917, "num_tokens": 547756907.0, "step": 14363 }, { "epoch": 1.8272484416740873, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.987133026123047, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8658611178398132, "num_tokens": 547793545.0, "step": 14364 }, { "epoch": 1.8273756519526778, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.049371719360352, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8606158494949341, "num_tokens": 547836845.0, "step": 14365 }, { "epoch": 1.8275028622312681, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.051446914672852, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8761717677116394, "num_tokens": 547875849.0, "step": 14366 }, { "epoch": 1.8276300725098586, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.025864601135254, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8640552759170532, "num_tokens": 547918672.0, "step": 14367 }, { "epoch": 1.8277572827884492, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.14713191986084, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8578326106071472, "num_tokens": 547958681.0, "step": 14368 }, { "epoch": 1.8278844930670397, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.988300323486328, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8647438287734985, "num_tokens": 547996940.0, "step": 14369 }, { "epoch": 1.8280117033456302, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.0997953414917, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8782640099525452, "num_tokens": 548030188.0, "step": 14370 }, { "epoch": 1.8281389136242208, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.0244140625, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8858605623245239, "num_tokens": 548066891.0, "step": 14371 }, { "epoch": 1.8282661239028113, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.055776596069336, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8753728270530701, "num_tokens": 548104664.0, "step": 14372 }, { "epoch": 1.8283933341814018, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.011798858642578, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8661744594573975, "num_tokens": 548143938.0, "step": 14373 }, { "epoch": 1.8285205444599923, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.110825538635254, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8690457344055176, "num_tokens": 548183213.0, "step": 14374 }, { "epoch": 1.8286477547385829, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.099570274353027, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8542441129684448, "num_tokens": 548221304.0, "step": 14375 }, { "epoch": 1.8287749650171734, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.0508451461792, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.867749810218811, "num_tokens": 548261123.0, "step": 14376 }, { "epoch": 1.828902175295764, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.068181991577148, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8504290580749512, "num_tokens": 548305040.0, "step": 14377 }, { "epoch": 1.8290293855743545, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.103227615356445, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8771491646766663, "num_tokens": 548338538.0, "step": 14378 }, { "epoch": 1.829156595852945, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.045225143432617, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8732746839523315, "num_tokens": 548378510.0, "step": 14379 }, { "epoch": 1.8292838061315355, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.094675064086914, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8641555905342102, "num_tokens": 548415790.0, "step": 14380 }, { "epoch": 1.829411016410126, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.084325790405273, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8609380722045898, "num_tokens": 548447294.0, "step": 14381 }, { "epoch": 1.8295382266887166, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.098276138305664, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8768022060394287, "num_tokens": 548484617.0, "step": 14382 }, { "epoch": 1.829665436967307, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.01755428314209, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.866478443145752, "num_tokens": 548527552.0, "step": 14383 }, { "epoch": 1.8297926472458976, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.076547622680664, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8887518048286438, "num_tokens": 548561322.0, "step": 14384 }, { "epoch": 1.829919857524488, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.10195541381836, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8795254826545715, "num_tokens": 548594810.0, "step": 14385 }, { "epoch": 1.8300470678030785, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.016096115112305, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8635284304618835, "num_tokens": 548631094.0, "step": 14386 }, { "epoch": 1.830174278081669, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.095636367797852, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8720384240150452, "num_tokens": 548669162.0, "step": 14387 }, { "epoch": 1.8303014883602595, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.086141586303711, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8694937229156494, "num_tokens": 548704555.0, "step": 14388 }, { "epoch": 1.83042869863885, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.105804443359375, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8838551640510559, "num_tokens": 548741852.0, "step": 14389 }, { "epoch": 1.8305559089174406, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.993634223937988, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8586315512657166, "num_tokens": 548788455.0, "step": 14390 }, { "epoch": 1.8306831191960309, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.058156967163086, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8773613572120667, "num_tokens": 548826347.0, "step": 14391 }, { "epoch": 1.8308103294746214, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.044282913208008, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8635926246643066, "num_tokens": 548865428.0, "step": 14392 }, { "epoch": 1.830937539753212, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.100239753723145, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8579808473587036, "num_tokens": 548901671.0, "step": 14393 }, { "epoch": 1.8310647500318025, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.001486778259277, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8663580417633057, "num_tokens": 548939234.0, "step": 14394 }, { "epoch": 1.831191960310393, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.119085311889648, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8749600648880005, "num_tokens": 548975334.0, "step": 14395 }, { "epoch": 1.8313191705889835, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.007131576538086, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8638055324554443, "num_tokens": 549016372.0, "step": 14396 }, { "epoch": 1.831446380867574, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.03172492980957, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.877295970916748, "num_tokens": 549058049.0, "step": 14397 }, { "epoch": 1.8315735911461646, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.07455062866211, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8577483296394348, "num_tokens": 549097855.0, "step": 14398 }, { "epoch": 1.831700801424755, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.021943092346191, "learning_rate": 1e-06, "loss": 0.3409, "mean_token_accuracy": 0.88966965675354, "num_tokens": 549130906.0, "step": 14399 }, { "epoch": 1.8318280117033456, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.10297966003418, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8514742851257324, "num_tokens": 549176574.0, "step": 14400 }, { "epoch": 1.8319552219819362, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.046586036682129, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8684097528457642, "num_tokens": 549213945.0, "step": 14401 }, { "epoch": 1.8320824322605267, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.057318687438965, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8614756464958191, "num_tokens": 549255843.0, "step": 14402 }, { "epoch": 1.8322096425391172, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.015251159667969, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8745274543762207, "num_tokens": 549292573.0, "step": 14403 }, { "epoch": 1.8323368528177078, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.068886756896973, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8832651376724243, "num_tokens": 549335291.0, "step": 14404 }, { "epoch": 1.8324640630962983, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.036943435668945, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.857367753982544, "num_tokens": 549381337.0, "step": 14405 }, { "epoch": 1.8325912733748888, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.07065200805664, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8810814619064331, "num_tokens": 549418180.0, "step": 14406 }, { "epoch": 1.8327184836534793, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.984057426452637, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8776261806488037, "num_tokens": 549461623.0, "step": 14407 }, { "epoch": 1.8328456939320699, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.036498069763184, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8733826875686646, "num_tokens": 549500791.0, "step": 14408 }, { "epoch": 1.8329729042106602, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.043252944946289, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8767608404159546, "num_tokens": 549533343.0, "step": 14409 }, { "epoch": 1.8331001144892507, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.107659339904785, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8619648218154907, "num_tokens": 549572707.0, "step": 14410 }, { "epoch": 1.8332273247678412, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.995816230773926, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8766794204711914, "num_tokens": 549609772.0, "step": 14411 }, { "epoch": 1.8333545350464318, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.031424522399902, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8589920997619629, "num_tokens": 549648626.0, "step": 14412 }, { "epoch": 1.8334817453250223, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.07712173461914, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8585190176963806, "num_tokens": 549693211.0, "step": 14413 }, { "epoch": 1.8336089556036128, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.02476978302002, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8637527227401733, "num_tokens": 549729463.0, "step": 14414 }, { "epoch": 1.8337361658822031, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.05235767364502, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8753765225410461, "num_tokens": 549769139.0, "step": 14415 }, { "epoch": 1.8338633761607936, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.07402515411377, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8732392191886902, "num_tokens": 549807067.0, "step": 14416 }, { "epoch": 1.8339905864393842, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.08389949798584, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8690395951271057, "num_tokens": 549842689.0, "step": 14417 }, { "epoch": 1.8341177967179747, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.07563591003418, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8896392583847046, "num_tokens": 549876108.0, "step": 14418 }, { "epoch": 1.8342450069965652, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.101093292236328, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8840245008468628, "num_tokens": 549909656.0, "step": 14419 }, { "epoch": 1.8343722172751558, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.027871131896973, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8621339797973633, "num_tokens": 549949156.0, "step": 14420 }, { "epoch": 1.8344994275537463, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.068428039550781, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8774468302726746, "num_tokens": 549990256.0, "step": 14421 }, { "epoch": 1.8346266378323368, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.070343017578125, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8793456554412842, "num_tokens": 550023557.0, "step": 14422 }, { "epoch": 1.8347538481109273, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.127449035644531, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8670665621757507, "num_tokens": 550069402.0, "step": 14423 }, { "epoch": 1.8348810583895179, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.99105453491211, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8775968551635742, "num_tokens": 550110860.0, "step": 14424 }, { "epoch": 1.8350082686681084, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.100346565246582, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8549103736877441, "num_tokens": 550151736.0, "step": 14425 }, { "epoch": 1.835135478946699, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.070948600769043, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8731765747070312, "num_tokens": 550187731.0, "step": 14426 }, { "epoch": 1.8352626892252895, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.065739631652832, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8723817467689514, "num_tokens": 550220545.0, "step": 14427 }, { "epoch": 1.83538989950388, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.165339469909668, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8746784329414368, "num_tokens": 550254755.0, "step": 14428 }, { "epoch": 1.8355171097824705, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.115777969360352, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8768075704574585, "num_tokens": 550295191.0, "step": 14429 }, { "epoch": 1.835644320061061, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.075715065002441, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8649078607559204, "num_tokens": 550332984.0, "step": 14430 }, { "epoch": 1.8357715303396516, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.097427368164062, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8726751804351807, "num_tokens": 550377470.0, "step": 14431 }, { "epoch": 1.835898740618242, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.081435203552246, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8740501999855042, "num_tokens": 550416123.0, "step": 14432 }, { "epoch": 1.8360259508968326, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.095159530639648, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8754194974899292, "num_tokens": 550458062.0, "step": 14433 }, { "epoch": 1.836153161175423, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.094709396362305, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.86378413438797, "num_tokens": 550496485.0, "step": 14434 }, { "epoch": 1.8362803714540135, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.079070091247559, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8721660375595093, "num_tokens": 550536500.0, "step": 14435 }, { "epoch": 1.836407581732604, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.076692581176758, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8646367192268372, "num_tokens": 550575611.0, "step": 14436 }, { "epoch": 1.8365347920111945, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.151556968688965, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.862747073173523, "num_tokens": 550616431.0, "step": 14437 }, { "epoch": 1.836662002289785, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.13913345336914, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8806734681129456, "num_tokens": 550653626.0, "step": 14438 }, { "epoch": 1.8367892125683756, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.117721557617188, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8669092655181885, "num_tokens": 550690899.0, "step": 14439 }, { "epoch": 1.8369164228469659, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.137883186340332, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8710535764694214, "num_tokens": 550734408.0, "step": 14440 }, { "epoch": 1.8370436331255564, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.04653549194336, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8581564426422119, "num_tokens": 550767888.0, "step": 14441 }, { "epoch": 1.837170843404147, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.161762237548828, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.871206521987915, "num_tokens": 550807192.0, "step": 14442 }, { "epoch": 1.8372980536827375, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.095752716064453, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8789318203926086, "num_tokens": 550848006.0, "step": 14443 }, { "epoch": 1.837425263961328, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.102130889892578, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.870503842830658, "num_tokens": 550888350.0, "step": 14444 }, { "epoch": 1.8375524742399185, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.086943626403809, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8768881559371948, "num_tokens": 550925389.0, "step": 14445 }, { "epoch": 1.837679684518509, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.13889217376709, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8749306797981262, "num_tokens": 550962610.0, "step": 14446 }, { "epoch": 1.8378068947970996, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.038301467895508, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8664368391036987, "num_tokens": 551001018.0, "step": 14447 }, { "epoch": 1.83793410507569, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.126886367797852, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8720135688781738, "num_tokens": 551038505.0, "step": 14448 }, { "epoch": 1.8380613153542806, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.121630668640137, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8817237019538879, "num_tokens": 551072235.0, "step": 14449 }, { "epoch": 1.8381885256328712, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.100671768188477, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8837518692016602, "num_tokens": 551110129.0, "step": 14450 }, { "epoch": 1.8383157359114617, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.99691104888916, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.872105062007904, "num_tokens": 551152380.0, "step": 14451 }, { "epoch": 1.8384429461900522, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.088763236999512, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8815151453018188, "num_tokens": 551195942.0, "step": 14452 }, { "epoch": 1.8385701564686427, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.044731140136719, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8632354736328125, "num_tokens": 551238502.0, "step": 14453 }, { "epoch": 1.8386973667472333, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.134800910949707, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8809669017791748, "num_tokens": 551278091.0, "step": 14454 }, { "epoch": 1.8388245770258238, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.068916320800781, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.851222813129425, "num_tokens": 551314940.0, "step": 14455 }, { "epoch": 1.8389517873044143, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.078105926513672, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8581724166870117, "num_tokens": 551356427.0, "step": 14456 }, { "epoch": 1.8390789975830049, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.029496192932129, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8754200339317322, "num_tokens": 551400436.0, "step": 14457 }, { "epoch": 1.8392062078615952, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.093974113464355, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.869541347026825, "num_tokens": 551438335.0, "step": 14458 }, { "epoch": 1.8393334181401857, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.098583221435547, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8740422129631042, "num_tokens": 551471214.0, "step": 14459 }, { "epoch": 1.8394606284187762, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.07774543762207, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8577775955200195, "num_tokens": 551504071.0, "step": 14460 }, { "epoch": 1.8395878386973668, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.080120086669922, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.865456223487854, "num_tokens": 551538456.0, "step": 14461 }, { "epoch": 1.8397150489759573, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.033221244812012, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8747299909591675, "num_tokens": 551575334.0, "step": 14462 }, { "epoch": 1.8398422592545478, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.001209259033203, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8795737028121948, "num_tokens": 551617127.0, "step": 14463 }, { "epoch": 1.8399694695331381, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.095693588256836, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8753288984298706, "num_tokens": 551655548.0, "step": 14464 }, { "epoch": 1.8400966798117286, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.108180046081543, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8677586317062378, "num_tokens": 551691807.0, "step": 14465 }, { "epoch": 1.8402238900903192, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.052753448486328, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8791097402572632, "num_tokens": 551729805.0, "step": 14466 }, { "epoch": 1.8403511003689097, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.04517936706543, "learning_rate": 1e-06, "loss": 0.4868, "mean_token_accuracy": 0.8481521606445312, "num_tokens": 551775220.0, "step": 14467 }, { "epoch": 1.8404783106475002, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.112801551818848, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8624439239501953, "num_tokens": 551816242.0, "step": 14468 }, { "epoch": 1.8406055209260908, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.044273376464844, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8591800928115845, "num_tokens": 551854929.0, "step": 14469 }, { "epoch": 1.8407327312046813, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.047157287597656, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8688498735427856, "num_tokens": 551898877.0, "step": 14470 }, { "epoch": 1.8408599414832718, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.076321601867676, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8870232105255127, "num_tokens": 551934252.0, "step": 14471 }, { "epoch": 1.8409871517618623, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.045117378234863, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8820569515228271, "num_tokens": 551974779.0, "step": 14472 }, { "epoch": 1.8411143620404529, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.737366676330566, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8703626990318298, "num_tokens": 552011627.0, "step": 14473 }, { "epoch": 1.8412415723190434, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.032312393188477, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8501005172729492, "num_tokens": 552045333.0, "step": 14474 }, { "epoch": 1.841368782597634, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.952465057373047, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8645904660224915, "num_tokens": 552079974.0, "step": 14475 }, { "epoch": 1.8414959928762245, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.046225547790527, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8784177303314209, "num_tokens": 552117507.0, "step": 14476 }, { "epoch": 1.841623203154815, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.036438941955566, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.870488166809082, "num_tokens": 552156009.0, "step": 14477 }, { "epoch": 1.8417504134334055, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.043083190917969, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8793042898178101, "num_tokens": 552189092.0, "step": 14478 }, { "epoch": 1.841877623711996, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.093788146972656, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8692370653152466, "num_tokens": 552228305.0, "step": 14479 }, { "epoch": 1.8420048339905866, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.105470657348633, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8718322515487671, "num_tokens": 552269402.0, "step": 14480 }, { "epoch": 1.842132044269177, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.144944190979004, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8674043416976929, "num_tokens": 552304870.0, "step": 14481 }, { "epoch": 1.8422592545477676, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.10867691040039, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8670824766159058, "num_tokens": 552336802.0, "step": 14482 }, { "epoch": 1.842386464826358, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.112820625305176, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8829091191291809, "num_tokens": 552376124.0, "step": 14483 }, { "epoch": 1.8425136751049485, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.072356224060059, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8611692190170288, "num_tokens": 552412285.0, "step": 14484 }, { "epoch": 1.842640885383539, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.993870735168457, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8720543384552002, "num_tokens": 552447527.0, "step": 14485 }, { "epoch": 1.8427680956621295, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.070202827453613, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8731687068939209, "num_tokens": 552486569.0, "step": 14486 }, { "epoch": 1.84289530594072, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.106805801391602, "learning_rate": 1e-06, "loss": 0.5405, "mean_token_accuracy": 0.8359958529472351, "num_tokens": 552529158.0, "step": 14487 }, { "epoch": 1.8430225162193106, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.043891906738281, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8685571551322937, "num_tokens": 552572183.0, "step": 14488 }, { "epoch": 1.8431497264979009, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.03439998626709, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8678855895996094, "num_tokens": 552609785.0, "step": 14489 }, { "epoch": 1.8432769367764914, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.065018653869629, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8765251636505127, "num_tokens": 552641243.0, "step": 14490 }, { "epoch": 1.843404147055082, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.08249568939209, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8623962998390198, "num_tokens": 552677595.0, "step": 14491 }, { "epoch": 1.8435313573336725, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.091106414794922, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8714245557785034, "num_tokens": 552715492.0, "step": 14492 }, { "epoch": 1.843658567612263, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.020004272460938, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8693158626556396, "num_tokens": 552757410.0, "step": 14493 }, { "epoch": 1.8437857778908535, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.037849426269531, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.886078417301178, "num_tokens": 552793794.0, "step": 14494 }, { "epoch": 1.843912988169444, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.049766540527344, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8554949760437012, "num_tokens": 552836010.0, "step": 14495 }, { "epoch": 1.8440401984480346, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.130406379699707, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.872576892375946, "num_tokens": 552874136.0, "step": 14496 }, { "epoch": 1.844167408726625, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.086494445800781, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8671560287475586, "num_tokens": 552912314.0, "step": 14497 }, { "epoch": 1.8442946190052156, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.067743301391602, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8579124808311462, "num_tokens": 552953998.0, "step": 14498 }, { "epoch": 1.8444218292838062, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.061161041259766, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8885824680328369, "num_tokens": 552987310.0, "step": 14499 }, { "epoch": 1.8445490395623967, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.142287254333496, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.868581235408783, "num_tokens": 553019830.0, "step": 14500 }, { "epoch": 1.8446762498409872, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.93556022644043, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8695598244667053, "num_tokens": 553066695.0, "step": 14501 }, { "epoch": 1.8448034601195777, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.155936241149902, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8619111776351929, "num_tokens": 553098943.0, "step": 14502 }, { "epoch": 1.8449306703981683, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.011201858520508, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8748867511749268, "num_tokens": 553136106.0, "step": 14503 }, { "epoch": 1.8450578806767588, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.072650909423828, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8768531084060669, "num_tokens": 553170014.0, "step": 14504 }, { "epoch": 1.8451850909553493, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.045841217041016, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8648889660835266, "num_tokens": 553213297.0, "step": 14505 }, { "epoch": 1.8453123012339399, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.056175231933594, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8716376423835754, "num_tokens": 553248559.0, "step": 14506 }, { "epoch": 1.8454395115125302, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.118687629699707, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8718295693397522, "num_tokens": 553289140.0, "step": 14507 }, { "epoch": 1.8455667217911207, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.052282333374023, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8627462387084961, "num_tokens": 553330619.0, "step": 14508 }, { "epoch": 1.8456939320697112, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.089607238769531, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8818747997283936, "num_tokens": 553371295.0, "step": 14509 }, { "epoch": 1.8458211423483017, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.99606990814209, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8666204214096069, "num_tokens": 553410237.0, "step": 14510 }, { "epoch": 1.8459483526268923, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.119617462158203, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8775574564933777, "num_tokens": 553446727.0, "step": 14511 }, { "epoch": 1.8460755629054828, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.036076545715332, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8766987919807434, "num_tokens": 553488626.0, "step": 14512 }, { "epoch": 1.846202773184073, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.074106216430664, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8634907603263855, "num_tokens": 553528294.0, "step": 14513 }, { "epoch": 1.8463299834626636, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.025965690612793, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8742687702178955, "num_tokens": 553563738.0, "step": 14514 }, { "epoch": 1.8464571937412542, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.087735176086426, "learning_rate": 1e-06, "loss": 0.5166, "mean_token_accuracy": 0.8382912278175354, "num_tokens": 553603351.0, "step": 14515 }, { "epoch": 1.8465844040198447, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.085482597351074, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8683093786239624, "num_tokens": 553643655.0, "step": 14516 }, { "epoch": 1.8467116142984352, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.061920166015625, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8824026584625244, "num_tokens": 553677944.0, "step": 14517 }, { "epoch": 1.8468388245770258, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.111504554748535, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8660770654678345, "num_tokens": 553712120.0, "step": 14518 }, { "epoch": 1.8469660348556163, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.069244384765625, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8736647367477417, "num_tokens": 553751442.0, "step": 14519 }, { "epoch": 1.8470932451342068, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.087109565734863, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8762850761413574, "num_tokens": 553782924.0, "step": 14520 }, { "epoch": 1.8472204554127973, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.112395286560059, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8613288402557373, "num_tokens": 553818861.0, "step": 14521 }, { "epoch": 1.8473476656913879, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.078042030334473, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8696006536483765, "num_tokens": 553855593.0, "step": 14522 }, { "epoch": 1.8474748759699784, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.138290405273438, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8687342405319214, "num_tokens": 553896028.0, "step": 14523 }, { "epoch": 1.847602086248569, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.045215606689453, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8742501735687256, "num_tokens": 553935212.0, "step": 14524 }, { "epoch": 1.8477292965271594, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.075447082519531, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8695023059844971, "num_tokens": 553971231.0, "step": 14525 }, { "epoch": 1.84785650680575, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.067156791687012, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8792446851730347, "num_tokens": 554006782.0, "step": 14526 }, { "epoch": 1.8479837170843405, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.129356384277344, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8630069494247437, "num_tokens": 554043427.0, "step": 14527 }, { "epoch": 1.848110927362931, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.020630836486816, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.860008716583252, "num_tokens": 554084187.0, "step": 14528 }, { "epoch": 1.8482381376415216, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.162522315979004, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8705476522445679, "num_tokens": 554122186.0, "step": 14529 }, { "epoch": 1.848365347920112, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.08198356628418, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8700354099273682, "num_tokens": 554161388.0, "step": 14530 }, { "epoch": 1.8484925581987026, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.087658882141113, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8623644113540649, "num_tokens": 554199837.0, "step": 14531 }, { "epoch": 1.848619768477293, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.140015602111816, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8613512516021729, "num_tokens": 554239567.0, "step": 14532 }, { "epoch": 1.8487469787558835, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.979891777038574, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8736765384674072, "num_tokens": 554277040.0, "step": 14533 }, { "epoch": 1.848874189034474, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.070287704467773, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8736212253570557, "num_tokens": 554318994.0, "step": 14534 }, { "epoch": 1.8490013993130645, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.132935523986816, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8641523122787476, "num_tokens": 554355334.0, "step": 14535 }, { "epoch": 1.849128609591655, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.114154815673828, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8593775629997253, "num_tokens": 554396504.0, "step": 14536 }, { "epoch": 1.8492558198702456, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.09408950805664, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8767005801200867, "num_tokens": 554429772.0, "step": 14537 }, { "epoch": 1.8493830301488359, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.02860164642334, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8725178837776184, "num_tokens": 554470102.0, "step": 14538 }, { "epoch": 1.8495102404274264, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.072802543640137, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8622164726257324, "num_tokens": 554511632.0, "step": 14539 }, { "epoch": 1.849637450706017, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.12132453918457, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8665738105773926, "num_tokens": 554553028.0, "step": 14540 }, { "epoch": 1.8497646609846075, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.047036170959473, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.864054799079895, "num_tokens": 554585109.0, "step": 14541 }, { "epoch": 1.849891871263198, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.102545738220215, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8711907863616943, "num_tokens": 554621141.0, "step": 14542 }, { "epoch": 1.8500190815417885, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.006636619567871, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8690329790115356, "num_tokens": 554660838.0, "step": 14543 }, { "epoch": 1.850146291820379, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.198027610778809, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8719488978385925, "num_tokens": 554696338.0, "step": 14544 }, { "epoch": 1.8502735020989696, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.099092483520508, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8820064067840576, "num_tokens": 554729782.0, "step": 14545 }, { "epoch": 1.85040071237756, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 13.065001487731934, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8856239318847656, "num_tokens": 554767566.0, "step": 14546 }, { "epoch": 1.8505279226561506, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.113418579101562, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8754464387893677, "num_tokens": 554806063.0, "step": 14547 }, { "epoch": 1.8506551329347412, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.126240730285645, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8672024011611938, "num_tokens": 554839418.0, "step": 14548 }, { "epoch": 1.8507823432133317, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.097105979919434, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8658055663108826, "num_tokens": 554881888.0, "step": 14549 }, { "epoch": 1.8509095534919222, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.05134105682373, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8880784511566162, "num_tokens": 554921561.0, "step": 14550 }, { "epoch": 1.8510367637705127, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.133216857910156, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8465045094490051, "num_tokens": 554954380.0, "step": 14551 }, { "epoch": 1.8511639740491033, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.0438814163208, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.871228814125061, "num_tokens": 554994781.0, "step": 14552 }, { "epoch": 1.8512911843276938, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.109829902648926, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8791706562042236, "num_tokens": 555024938.0, "step": 14553 }, { "epoch": 1.8514183946062843, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.072361946105957, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8584154844284058, "num_tokens": 555059785.0, "step": 14554 }, { "epoch": 1.8515456048848749, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.02752685546875, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8445917367935181, "num_tokens": 555097502.0, "step": 14555 }, { "epoch": 1.8516728151634652, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.043444633483887, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8683110475540161, "num_tokens": 555139976.0, "step": 14556 }, { "epoch": 1.8518000254420557, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.114653587341309, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8730823397636414, "num_tokens": 555181250.0, "step": 14557 }, { "epoch": 1.8519272357206462, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.082884788513184, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8513938188552856, "num_tokens": 555214927.0, "step": 14558 }, { "epoch": 1.8520544459992367, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.056683540344238, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8665515184402466, "num_tokens": 555255795.0, "step": 14559 }, { "epoch": 1.8521816562778273, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.049121856689453, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8731858730316162, "num_tokens": 555294177.0, "step": 14560 }, { "epoch": 1.8523088665564178, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.126300811767578, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.882788896560669, "num_tokens": 555331548.0, "step": 14561 }, { "epoch": 1.852436076835008, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.051592826843262, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8818231821060181, "num_tokens": 555368426.0, "step": 14562 }, { "epoch": 1.8525632871135986, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.08860969543457, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8867248892784119, "num_tokens": 555411594.0, "step": 14563 }, { "epoch": 1.8526904973921892, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.095755577087402, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.862608015537262, "num_tokens": 555450412.0, "step": 14564 }, { "epoch": 1.8528177076707797, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.060948371887207, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8784558773040771, "num_tokens": 555489067.0, "step": 14565 }, { "epoch": 1.8529449179493702, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.101153373718262, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8761700987815857, "num_tokens": 555524552.0, "step": 14566 }, { "epoch": 1.8530721282279607, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.078866958618164, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8436861634254456, "num_tokens": 555563766.0, "step": 14567 }, { "epoch": 1.8531993385065513, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.05019760131836, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8766388297080994, "num_tokens": 555601410.0, "step": 14568 }, { "epoch": 1.8533265487851418, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.081892013549805, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.870207667350769, "num_tokens": 555641346.0, "step": 14569 }, { "epoch": 1.8534537590637323, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.090405464172363, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8805453777313232, "num_tokens": 555676537.0, "step": 14570 }, { "epoch": 1.8535809693423229, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.109358787536621, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8555041551589966, "num_tokens": 555721634.0, "step": 14571 }, { "epoch": 1.8537081796209134, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.98929214477539, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8698059320449829, "num_tokens": 555765447.0, "step": 14572 }, { "epoch": 1.853835389899504, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.09758472442627, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8700609803199768, "num_tokens": 555798598.0, "step": 14573 }, { "epoch": 1.8539626001780944, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.105899810791016, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8799241781234741, "num_tokens": 555829465.0, "step": 14574 }, { "epoch": 1.854089810456685, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.037442207336426, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8717250823974609, "num_tokens": 555866577.0, "step": 14575 }, { "epoch": 1.8542170207352755, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.098306655883789, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.873724102973938, "num_tokens": 555900472.0, "step": 14576 }, { "epoch": 1.854344231013866, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.096654891967773, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8660573959350586, "num_tokens": 555936356.0, "step": 14577 }, { "epoch": 1.8544714412924566, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.063020706176758, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8639125823974609, "num_tokens": 555975338.0, "step": 14578 }, { "epoch": 1.854598651571047, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.133191108703613, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8597134351730347, "num_tokens": 556019651.0, "step": 14579 }, { "epoch": 1.8547258618496376, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.0919189453125, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8661100268363953, "num_tokens": 556054530.0, "step": 14580 }, { "epoch": 1.854853072128228, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.143601417541504, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8664474487304688, "num_tokens": 556091985.0, "step": 14581 }, { "epoch": 1.8549802824068184, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.043502807617188, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8725115060806274, "num_tokens": 556128777.0, "step": 14582 }, { "epoch": 1.855107492685409, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.152313232421875, "learning_rate": 1e-06, "loss": 0.3463, "mean_token_accuracy": 0.8893142938613892, "num_tokens": 556164286.0, "step": 14583 }, { "epoch": 1.8552347029639995, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.088139533996582, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8776509761810303, "num_tokens": 556199141.0, "step": 14584 }, { "epoch": 1.85536191324259, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.068427085876465, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.875381350517273, "num_tokens": 556237104.0, "step": 14585 }, { "epoch": 1.8554891235211806, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.096550941467285, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8594872951507568, "num_tokens": 556275343.0, "step": 14586 }, { "epoch": 1.8556163337997709, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.09793472290039, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.885764479637146, "num_tokens": 556316539.0, "step": 14587 }, { "epoch": 1.8557435440783614, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.06636905670166, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8834359645843506, "num_tokens": 556351101.0, "step": 14588 }, { "epoch": 1.855870754356952, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.101554870605469, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8832722902297974, "num_tokens": 556382458.0, "step": 14589 }, { "epoch": 1.8559979646355425, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.101764678955078, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8729931116104126, "num_tokens": 556416980.0, "step": 14590 }, { "epoch": 1.856125174914133, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.130932807922363, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8770624399185181, "num_tokens": 556451868.0, "step": 14591 }, { "epoch": 1.8562523851927235, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.131866455078125, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.860187292098999, "num_tokens": 556485682.0, "step": 14592 }, { "epoch": 1.856379595471314, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.110426902770996, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8821723461151123, "num_tokens": 556524857.0, "step": 14593 }, { "epoch": 1.8565068057499046, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.093092918395996, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8751095533370972, "num_tokens": 556565763.0, "step": 14594 }, { "epoch": 1.856634016028495, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.122584342956543, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.859484076499939, "num_tokens": 556606202.0, "step": 14595 }, { "epoch": 1.8567612263070856, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.11282730102539, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8795472979545593, "num_tokens": 556646906.0, "step": 14596 }, { "epoch": 1.8568884365856761, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.063493728637695, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8726727962493896, "num_tokens": 556684249.0, "step": 14597 }, { "epoch": 1.8570156468642667, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.118531227111816, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8601278066635132, "num_tokens": 556726840.0, "step": 14598 }, { "epoch": 1.8571428571428572, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.166419982910156, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8737854361534119, "num_tokens": 556764822.0, "step": 14599 }, { "epoch": 1.8572700674214477, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.124990463256836, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8688873052597046, "num_tokens": 556800884.0, "step": 14600 }, { "epoch": 1.8573972777000383, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.096965789794922, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.876666784286499, "num_tokens": 556841233.0, "step": 14601 }, { "epoch": 1.8575244879786288, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.13171100616455, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8647550940513611, "num_tokens": 556880995.0, "step": 14602 }, { "epoch": 1.8576516982572193, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.133641242980957, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8758174777030945, "num_tokens": 556914367.0, "step": 14603 }, { "epoch": 1.8577789085358098, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.193681716918945, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8508415222167969, "num_tokens": 556955944.0, "step": 14604 }, { "epoch": 1.8579061188144002, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.11472225189209, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8538042306900024, "num_tokens": 556989851.0, "step": 14605 }, { "epoch": 1.8580333290929907, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.12087631225586, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8700714111328125, "num_tokens": 557033662.0, "step": 14606 }, { "epoch": 1.8581605393715812, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.069107055664062, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8636217713356018, "num_tokens": 557074869.0, "step": 14607 }, { "epoch": 1.8582877496501717, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.08049201965332, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8599672317504883, "num_tokens": 557110164.0, "step": 14608 }, { "epoch": 1.8584149599287623, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.069464683532715, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8708525896072388, "num_tokens": 557149309.0, "step": 14609 }, { "epoch": 1.8585421702073528, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.141035079956055, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8654496669769287, "num_tokens": 557180510.0, "step": 14610 }, { "epoch": 1.858669380485943, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.103266716003418, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8766795992851257, "num_tokens": 557218062.0, "step": 14611 }, { "epoch": 1.8587965907645336, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.980446815490723, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8629459738731384, "num_tokens": 557258963.0, "step": 14612 }, { "epoch": 1.8589238010431242, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.138764381408691, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8713893890380859, "num_tokens": 557296046.0, "step": 14613 }, { "epoch": 1.8590510113217147, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.068978309631348, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8680402040481567, "num_tokens": 557328395.0, "step": 14614 }, { "epoch": 1.8591782216003052, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.060688018798828, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8734890818595886, "num_tokens": 557363888.0, "step": 14615 }, { "epoch": 1.8593054318788957, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.13401985168457, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8629167079925537, "num_tokens": 557402944.0, "step": 14616 }, { "epoch": 1.8594326421574863, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.088361740112305, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8760761022567749, "num_tokens": 557442247.0, "step": 14617 }, { "epoch": 1.8595598524360768, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.06602668762207, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8648091554641724, "num_tokens": 557483580.0, "step": 14618 }, { "epoch": 1.8596870627146673, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.112667083740234, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8745230436325073, "num_tokens": 557520134.0, "step": 14619 }, { "epoch": 1.8598142729932579, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.09101390838623, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8652110695838928, "num_tokens": 557565726.0, "step": 14620 }, { "epoch": 1.8599414832718484, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.06466007232666, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8694081902503967, "num_tokens": 557602871.0, "step": 14621 }, { "epoch": 1.860068693550439, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.12775707244873, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8617439866065979, "num_tokens": 557640873.0, "step": 14622 }, { "epoch": 1.8601959038290294, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.079431533813477, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8638567924499512, "num_tokens": 557684233.0, "step": 14623 }, { "epoch": 1.86032311410762, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.105412483215332, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8746762275695801, "num_tokens": 557716895.0, "step": 14624 }, { "epoch": 1.8604503243862105, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.052545547485352, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8667893409729004, "num_tokens": 557758197.0, "step": 14625 }, { "epoch": 1.860577534664801, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.147347450256348, "learning_rate": 1e-06, "loss": 0.5061, "mean_token_accuracy": 0.843755304813385, "num_tokens": 557794196.0, "step": 14626 }, { "epoch": 1.8607047449433916, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.035879135131836, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8542736768722534, "num_tokens": 557833856.0, "step": 14627 }, { "epoch": 1.860831955221982, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.104930877685547, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8590645790100098, "num_tokens": 557866304.0, "step": 14628 }, { "epoch": 1.8609591655005726, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.056258201599121, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8827980160713196, "num_tokens": 557907065.0, "step": 14629 }, { "epoch": 1.861086375779163, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.058512687683105, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8657811880111694, "num_tokens": 557947326.0, "step": 14630 }, { "epoch": 1.8612135860577534, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.101012229919434, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8687947988510132, "num_tokens": 557990059.0, "step": 14631 }, { "epoch": 1.861340796336344, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.057241439819336, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8714295625686646, "num_tokens": 558028406.0, "step": 14632 }, { "epoch": 1.8614680066149345, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.160436630249023, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8656975030899048, "num_tokens": 558070239.0, "step": 14633 }, { "epoch": 1.861595216893525, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.09021282196045, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8732803463935852, "num_tokens": 558107685.0, "step": 14634 }, { "epoch": 1.8617224271721156, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.187783241271973, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8641977310180664, "num_tokens": 558143185.0, "step": 14635 }, { "epoch": 1.8618496374507059, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.09708023071289, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8783622980117798, "num_tokens": 558176207.0, "step": 14636 }, { "epoch": 1.8619768477292964, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.107575416564941, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8860540390014648, "num_tokens": 558212215.0, "step": 14637 }, { "epoch": 1.862104058007887, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.191999435424805, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8797338008880615, "num_tokens": 558248471.0, "step": 14638 }, { "epoch": 1.8622312682864774, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.099300384521484, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8575158715248108, "num_tokens": 558285211.0, "step": 14639 }, { "epoch": 1.862358478565068, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.168905258178711, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8635570406913757, "num_tokens": 558314434.0, "step": 14640 }, { "epoch": 1.8624856888436585, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.129794120788574, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8779431581497192, "num_tokens": 558353710.0, "step": 14641 }, { "epoch": 1.862612899122249, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.096816062927246, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8575630187988281, "num_tokens": 558390476.0, "step": 14642 }, { "epoch": 1.8627401094008396, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.04228401184082, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.879263162612915, "num_tokens": 558424917.0, "step": 14643 }, { "epoch": 1.86286731967943, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.149094581604004, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8643209338188171, "num_tokens": 558459831.0, "step": 14644 }, { "epoch": 1.8629945299580206, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.0154390335083, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8844724893569946, "num_tokens": 558498600.0, "step": 14645 }, { "epoch": 1.8631217402366111, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.123393058776855, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8607633709907532, "num_tokens": 558537465.0, "step": 14646 }, { "epoch": 1.8632489505152017, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.072049140930176, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8898122310638428, "num_tokens": 558575357.0, "step": 14647 }, { "epoch": 1.8633761607937922, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.02043342590332, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8675840497016907, "num_tokens": 558619422.0, "step": 14648 }, { "epoch": 1.8635033710723827, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.100587844848633, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8610714077949524, "num_tokens": 558657677.0, "step": 14649 }, { "epoch": 1.8636305813509733, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.099324226379395, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8781803250312805, "num_tokens": 558692912.0, "step": 14650 }, { "epoch": 1.8637577916295638, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.009821891784668, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8826056718826294, "num_tokens": 558733470.0, "step": 14651 }, { "epoch": 1.8638850019081543, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.116560935974121, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8666935563087463, "num_tokens": 558771899.0, "step": 14652 }, { "epoch": 1.8640122121867448, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.061040878295898, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8766674399375916, "num_tokens": 558813002.0, "step": 14653 }, { "epoch": 1.8641394224653351, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.12214183807373, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8548073768615723, "num_tokens": 558850827.0, "step": 14654 }, { "epoch": 1.8642666327439257, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.092388153076172, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8674148321151733, "num_tokens": 558889003.0, "step": 14655 }, { "epoch": 1.8643938430225162, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.153362274169922, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8768166303634644, "num_tokens": 558925849.0, "step": 14656 }, { "epoch": 1.8645210533011067, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.12006950378418, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8692183494567871, "num_tokens": 558967471.0, "step": 14657 }, { "epoch": 1.8646482635796973, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.048006057739258, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8733282089233398, "num_tokens": 559001938.0, "step": 14658 }, { "epoch": 1.8647754738582878, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.150949478149414, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8764832019805908, "num_tokens": 559040515.0, "step": 14659 }, { "epoch": 1.864902684136878, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.084572792053223, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8825560808181763, "num_tokens": 559077702.0, "step": 14660 }, { "epoch": 1.8650298944154686, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.134115219116211, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8632737994194031, "num_tokens": 559114741.0, "step": 14661 }, { "epoch": 1.8651571046940592, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.074210166931152, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8485466241836548, "num_tokens": 559155844.0, "step": 14662 }, { "epoch": 1.8652843149726497, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.050627708435059, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8721857070922852, "num_tokens": 559194964.0, "step": 14663 }, { "epoch": 1.8654115252512402, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.117447853088379, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8763343095779419, "num_tokens": 559230083.0, "step": 14664 }, { "epoch": 1.8655387355298307, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.0863037109375, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8811831474304199, "num_tokens": 559268397.0, "step": 14665 }, { "epoch": 1.8656659458084213, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.139327049255371, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8623791337013245, "num_tokens": 559302024.0, "step": 14666 }, { "epoch": 1.8657931560870118, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.132268905639648, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8682339191436768, "num_tokens": 559343192.0, "step": 14667 }, { "epoch": 1.8659203663656023, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.098010063171387, "learning_rate": 1e-06, "loss": 0.3386, "mean_token_accuracy": 0.8924286365509033, "num_tokens": 559380616.0, "step": 14668 }, { "epoch": 1.8660475766441929, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.102771759033203, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8807653188705444, "num_tokens": 559413621.0, "step": 14669 }, { "epoch": 1.8661747869227834, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.156925201416016, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8798348307609558, "num_tokens": 559447274.0, "step": 14670 }, { "epoch": 1.866301997201374, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.079309463500977, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8800505995750427, "num_tokens": 559484306.0, "step": 14671 }, { "epoch": 1.8664292074799644, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.04514217376709, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8768914937973022, "num_tokens": 559527126.0, "step": 14672 }, { "epoch": 1.866556417758555, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.051143646240234, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8752111196517944, "num_tokens": 559568342.0, "step": 14673 }, { "epoch": 1.8666836280371455, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.08507251739502, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.867892861366272, "num_tokens": 559609597.0, "step": 14674 }, { "epoch": 1.866810838315736, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.059554100036621, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8685824871063232, "num_tokens": 559646392.0, "step": 14675 }, { "epoch": 1.8669380485943265, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.137747764587402, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8741872310638428, "num_tokens": 559684836.0, "step": 14676 }, { "epoch": 1.867065258872917, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.089439392089844, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8746634721755981, "num_tokens": 559727793.0, "step": 14677 }, { "epoch": 1.8671924691515076, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.119817733764648, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8593921661376953, "num_tokens": 559768154.0, "step": 14678 }, { "epoch": 1.867319679430098, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.140923500061035, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8729079961776733, "num_tokens": 559798732.0, "step": 14679 }, { "epoch": 1.8674468897086884, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.07846450805664, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8694324493408203, "num_tokens": 559842796.0, "step": 14680 }, { "epoch": 1.867574099987279, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.096748352050781, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8675192594528198, "num_tokens": 559884198.0, "step": 14681 }, { "epoch": 1.8677013102658695, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.128952026367188, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8560208082199097, "num_tokens": 559916749.0, "step": 14682 }, { "epoch": 1.86782852054446, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.109943389892578, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8664432764053345, "num_tokens": 559958430.0, "step": 14683 }, { "epoch": 1.8679557308230506, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.152473449707031, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8873418569564819, "num_tokens": 559995720.0, "step": 14684 }, { "epoch": 1.8680829411016409, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.098257064819336, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8750296831130981, "num_tokens": 560033304.0, "step": 14685 }, { "epoch": 1.8682101513802314, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.141843795776367, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8692295551300049, "num_tokens": 560069879.0, "step": 14686 }, { "epoch": 1.868337361658822, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.135870933532715, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8691790699958801, "num_tokens": 560106853.0, "step": 14687 }, { "epoch": 1.8684645719374124, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.104497909545898, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8716484308242798, "num_tokens": 560144225.0, "step": 14688 }, { "epoch": 1.868591782216003, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.134552001953125, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8640555143356323, "num_tokens": 560186520.0, "step": 14689 }, { "epoch": 1.8687189924945935, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.062955856323242, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8736491799354553, "num_tokens": 560225458.0, "step": 14690 }, { "epoch": 1.868846202773184, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.162097930908203, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8754395246505737, "num_tokens": 560263401.0, "step": 14691 }, { "epoch": 1.8689734130517746, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.036136627197266, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.867088794708252, "num_tokens": 560305026.0, "step": 14692 }, { "epoch": 1.869100623330365, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.122879981994629, "learning_rate": 1e-06, "loss": 0.5062, "mean_token_accuracy": 0.8390054702758789, "num_tokens": 560354823.0, "step": 14693 }, { "epoch": 1.8692278336089556, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.088563919067383, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.885683000087738, "num_tokens": 560391888.0, "step": 14694 }, { "epoch": 1.8693550438875461, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.091758728027344, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8732457756996155, "num_tokens": 560425350.0, "step": 14695 }, { "epoch": 1.8694822541661367, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.13077449798584, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8814008235931396, "num_tokens": 560458521.0, "step": 14696 }, { "epoch": 1.8696094644447272, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.093839645385742, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8651095032691956, "num_tokens": 560506378.0, "step": 14697 }, { "epoch": 1.8697366747233177, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.150267601013184, "learning_rate": 1e-06, "loss": 0.3347, "mean_token_accuracy": 0.892535388469696, "num_tokens": 560544055.0, "step": 14698 }, { "epoch": 1.8698638850019083, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.122803688049316, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.880431056022644, "num_tokens": 560577240.0, "step": 14699 }, { "epoch": 1.8699910952804988, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.057795524597168, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8661179542541504, "num_tokens": 560615662.0, "step": 14700 }, { "epoch": 1.8701183055590893, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.05754280090332, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8659502863883972, "num_tokens": 560658386.0, "step": 14701 }, { "epoch": 1.8702455158376798, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.108834266662598, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8688119649887085, "num_tokens": 560696366.0, "step": 14702 }, { "epoch": 1.8703727261162701, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.126287460327148, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8593534827232361, "num_tokens": 560736419.0, "step": 14703 }, { "epoch": 1.8704999363948607, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.109682083129883, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8699551820755005, "num_tokens": 560777839.0, "step": 14704 }, { "epoch": 1.8706271466734512, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.093840599060059, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8804129362106323, "num_tokens": 560816167.0, "step": 14705 }, { "epoch": 1.8707543569520417, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.097210884094238, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8574767708778381, "num_tokens": 560854524.0, "step": 14706 }, { "epoch": 1.8708815672306323, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.149309158325195, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8677981495857239, "num_tokens": 560894393.0, "step": 14707 }, { "epoch": 1.8710087775092228, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.070695877075195, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8810666799545288, "num_tokens": 560938640.0, "step": 14708 }, { "epoch": 1.871135987787813, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.121707916259766, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8693878650665283, "num_tokens": 560976805.0, "step": 14709 }, { "epoch": 1.8712631980664036, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.125396728515625, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8672861456871033, "num_tokens": 561018940.0, "step": 14710 }, { "epoch": 1.8713904083449941, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.11064624786377, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8762034177780151, "num_tokens": 561065228.0, "step": 14711 }, { "epoch": 1.8715176186235847, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.08685302734375, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.862934947013855, "num_tokens": 561107540.0, "step": 14712 }, { "epoch": 1.8716448289021752, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.129121780395508, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8515514135360718, "num_tokens": 561144774.0, "step": 14713 }, { "epoch": 1.8717720391807657, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.153779029846191, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8511453866958618, "num_tokens": 561186774.0, "step": 14714 }, { "epoch": 1.8718992494593563, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.083066940307617, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8836773633956909, "num_tokens": 561223997.0, "step": 14715 }, { "epoch": 1.8720264597379468, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.124256134033203, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.878220796585083, "num_tokens": 561261583.0, "step": 14716 }, { "epoch": 1.8721536700165373, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.07578182220459, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8755441904067993, "num_tokens": 561302681.0, "step": 14717 }, { "epoch": 1.8722808802951278, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.12061882019043, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8694252967834473, "num_tokens": 561338402.0, "step": 14718 }, { "epoch": 1.8724080905737184, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.067057609558105, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8745779991149902, "num_tokens": 561372892.0, "step": 14719 }, { "epoch": 1.872535300852309, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.142118453979492, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8675122261047363, "num_tokens": 561409474.0, "step": 14720 }, { "epoch": 1.8726625111308994, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.041132926940918, "learning_rate": 1e-06, "loss": 0.4853, "mean_token_accuracy": 0.8446444272994995, "num_tokens": 561450152.0, "step": 14721 }, { "epoch": 1.87278972140949, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.136626243591309, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.864852249622345, "num_tokens": 561489448.0, "step": 14722 }, { "epoch": 1.8729169316880805, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.113788604736328, "learning_rate": 1e-06, "loss": 0.4815, "mean_token_accuracy": 0.8481568694114685, "num_tokens": 561529684.0, "step": 14723 }, { "epoch": 1.873044141966671, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.114984512329102, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8719861507415771, "num_tokens": 561564259.0, "step": 14724 }, { "epoch": 1.8731713522452615, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.115763664245605, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.878223717212677, "num_tokens": 561609650.0, "step": 14725 }, { "epoch": 1.873298562523852, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.148847579956055, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8766215443611145, "num_tokens": 561645520.0, "step": 14726 }, { "epoch": 1.8734257728024426, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.057671546936035, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8755418062210083, "num_tokens": 561682771.0, "step": 14727 }, { "epoch": 1.873552983081033, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.102195739746094, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.860045313835144, "num_tokens": 561721317.0, "step": 14728 }, { "epoch": 1.8736801933596234, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.192124366760254, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8719450831413269, "num_tokens": 561763178.0, "step": 14729 }, { "epoch": 1.873807403638214, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.129073143005371, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8709126114845276, "num_tokens": 561802292.0, "step": 14730 }, { "epoch": 1.8739346139168045, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.143115997314453, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8788922429084778, "num_tokens": 561838610.0, "step": 14731 }, { "epoch": 1.874061824195395, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.134932518005371, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8744727969169617, "num_tokens": 561881992.0, "step": 14732 }, { "epoch": 1.8741890344739855, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.12033462524414, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8608160018920898, "num_tokens": 561917661.0, "step": 14733 }, { "epoch": 1.8743162447525759, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.123619079589844, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8719606399536133, "num_tokens": 561958218.0, "step": 14734 }, { "epoch": 1.8744434550311664, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.161869049072266, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8815317153930664, "num_tokens": 561993882.0, "step": 14735 }, { "epoch": 1.874570665309757, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.13623046875, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.871921956539154, "num_tokens": 562035443.0, "step": 14736 }, { "epoch": 1.8746978755883474, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.117815971374512, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.869110107421875, "num_tokens": 562072806.0, "step": 14737 }, { "epoch": 1.874825085866938, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.118228912353516, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8541373014450073, "num_tokens": 562113652.0, "step": 14738 }, { "epoch": 1.8749522961455285, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.151294708251953, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8624502420425415, "num_tokens": 562150600.0, "step": 14739 }, { "epoch": 1.875079506424119, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.68028450012207, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8720272183418274, "num_tokens": 562187730.0, "step": 14740 }, { "epoch": 1.8752067167027096, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.128911018371582, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.86993008852005, "num_tokens": 562228229.0, "step": 14741 }, { "epoch": 1.8753339269813, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 13.109993934631348, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8765155673027039, "num_tokens": 562262293.0, "step": 14742 }, { "epoch": 1.8754611372598906, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.13769817352295, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8679591417312622, "num_tokens": 562303389.0, "step": 14743 }, { "epoch": 1.8755883475384811, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.081218719482422, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8641373515129089, "num_tokens": 562344461.0, "step": 14744 }, { "epoch": 1.8757155578170717, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.183675765991211, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8529931306838989, "num_tokens": 562387068.0, "step": 14745 }, { "epoch": 1.8758427680956622, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.103843688964844, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8682407736778259, "num_tokens": 562428225.0, "step": 14746 }, { "epoch": 1.8759699783742527, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.088780403137207, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8707106709480286, "num_tokens": 562463879.0, "step": 14747 }, { "epoch": 1.8760971886528433, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.125603675842285, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8636155128479004, "num_tokens": 562502327.0, "step": 14748 }, { "epoch": 1.8762243989314338, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.13105297088623, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.875124454498291, "num_tokens": 562542162.0, "step": 14749 }, { "epoch": 1.8763516092100243, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.14959716796875, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8651168346405029, "num_tokens": 562583364.0, "step": 14750 }, { "epoch": 1.8764788194886148, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.141948699951172, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8724046349525452, "num_tokens": 562615870.0, "step": 14751 }, { "epoch": 1.8766060297672051, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.119388580322266, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8558982014656067, "num_tokens": 562655086.0, "step": 14752 }, { "epoch": 1.8767332400457957, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.16707706451416, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8625761866569519, "num_tokens": 562697532.0, "step": 14753 }, { "epoch": 1.8768604503243862, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.211819648742676, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8495700359344482, "num_tokens": 562731057.0, "step": 14754 }, { "epoch": 1.8769876606029767, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.175660133361816, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8592462539672852, "num_tokens": 562770402.0, "step": 14755 }, { "epoch": 1.8771148708815673, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.195232391357422, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8662891387939453, "num_tokens": 562806191.0, "step": 14756 }, { "epoch": 1.8772420811601578, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.14946174621582, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8683828115463257, "num_tokens": 562846864.0, "step": 14757 }, { "epoch": 1.877369291438748, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.100245475769043, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8647785782814026, "num_tokens": 562882887.0, "step": 14758 }, { "epoch": 1.8774965017173386, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.224752426147461, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8688371181488037, "num_tokens": 562919096.0, "step": 14759 }, { "epoch": 1.8776237119959291, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.125094413757324, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8438624143600464, "num_tokens": 562959336.0, "step": 14760 }, { "epoch": 1.8777509222745197, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.139883995056152, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8692089915275574, "num_tokens": 562997883.0, "step": 14761 }, { "epoch": 1.8778781325531102, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.125415802001953, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8791377544403076, "num_tokens": 563036763.0, "step": 14762 }, { "epoch": 1.8780053428317007, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.037639617919922, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8858827352523804, "num_tokens": 563076815.0, "step": 14763 }, { "epoch": 1.8781325531102913, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.175848007202148, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8803322315216064, "num_tokens": 563110511.0, "step": 14764 }, { "epoch": 1.8782597633888818, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.149438858032227, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8641986846923828, "num_tokens": 563140757.0, "step": 14765 }, { "epoch": 1.8783869736674723, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.10693359375, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8701404333114624, "num_tokens": 563179961.0, "step": 14766 }, { "epoch": 1.8785141839460628, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.071640014648438, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8770690560340881, "num_tokens": 563221355.0, "step": 14767 }, { "epoch": 1.8786413942246534, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.161917686462402, "learning_rate": 1e-06, "loss": 0.3427, "mean_token_accuracy": 0.891333818435669, "num_tokens": 563257427.0, "step": 14768 }, { "epoch": 1.878768604503244, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.17510986328125, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8670567870140076, "num_tokens": 563294371.0, "step": 14769 }, { "epoch": 1.8788958147818344, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.158062934875488, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8802287578582764, "num_tokens": 563338592.0, "step": 14770 }, { "epoch": 1.879023025060425, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.219712257385254, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8727079033851624, "num_tokens": 563373853.0, "step": 14771 }, { "epoch": 1.8791502353390155, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.145535469055176, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8696811199188232, "num_tokens": 563411137.0, "step": 14772 }, { "epoch": 1.879277445617606, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.167555809020996, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8735204935073853, "num_tokens": 563450294.0, "step": 14773 }, { "epoch": 1.8794046558961965, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.155888557434082, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8720209002494812, "num_tokens": 563496858.0, "step": 14774 }, { "epoch": 1.879531866174787, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.161972045898438, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8655486106872559, "num_tokens": 563539566.0, "step": 14775 }, { "epoch": 1.8796590764533776, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.171905517578125, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8785173892974854, "num_tokens": 563574363.0, "step": 14776 }, { "epoch": 1.879786286731968, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.191267013549805, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8719944953918457, "num_tokens": 563605540.0, "step": 14777 }, { "epoch": 1.8799134970105584, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.170445442199707, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8566657304763794, "num_tokens": 563646640.0, "step": 14778 }, { "epoch": 1.880040707289149, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.171215057373047, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8808216452598572, "num_tokens": 563683576.0, "step": 14779 }, { "epoch": 1.8801679175677395, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.150005340576172, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8628399968147278, "num_tokens": 563724088.0, "step": 14780 }, { "epoch": 1.88029512784633, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.186275482177734, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8691049814224243, "num_tokens": 563760897.0, "step": 14781 }, { "epoch": 1.8804223381249205, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.16071605682373, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8757801651954651, "num_tokens": 563800158.0, "step": 14782 }, { "epoch": 1.8805495484035109, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.176290512084961, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8755539655685425, "num_tokens": 563833288.0, "step": 14783 }, { "epoch": 1.8806767586821014, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.181082725524902, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8812161087989807, "num_tokens": 563874507.0, "step": 14784 }, { "epoch": 1.880803968960692, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.099297523498535, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8698083162307739, "num_tokens": 563912452.0, "step": 14785 }, { "epoch": 1.8809311792392824, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.220996856689453, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8530532121658325, "num_tokens": 563950565.0, "step": 14786 }, { "epoch": 1.881058389517873, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.181854248046875, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8607004880905151, "num_tokens": 563984924.0, "step": 14787 }, { "epoch": 1.8811855997964635, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.12659740447998, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.877066433429718, "num_tokens": 564022887.0, "step": 14788 }, { "epoch": 1.881312810075054, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.111833572387695, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8688759207725525, "num_tokens": 564064829.0, "step": 14789 }, { "epoch": 1.8814400203536445, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.178519248962402, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8485895991325378, "num_tokens": 564100858.0, "step": 14790 }, { "epoch": 1.881567230632235, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.123031616210938, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8636883497238159, "num_tokens": 564141477.0, "step": 14791 }, { "epoch": 1.8816944409108256, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.212176322937012, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.86326003074646, "num_tokens": 564179271.0, "step": 14792 }, { "epoch": 1.8818216511894161, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.133845329284668, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8757510185241699, "num_tokens": 564224336.0, "step": 14793 }, { "epoch": 1.8819488614680067, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.14716625213623, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.86448073387146, "num_tokens": 564265157.0, "step": 14794 }, { "epoch": 1.8820760717465972, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.230886459350586, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8760679960250854, "num_tokens": 564295023.0, "step": 14795 }, { "epoch": 1.8822032820251877, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.13487434387207, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8718634247779846, "num_tokens": 564333700.0, "step": 14796 }, { "epoch": 1.8823304923037782, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.201626777648926, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.856890082359314, "num_tokens": 564374342.0, "step": 14797 }, { "epoch": 1.8824577025823688, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.143413543701172, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.871505618095398, "num_tokens": 564422877.0, "step": 14798 }, { "epoch": 1.8825849128609593, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.097759246826172, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8698814511299133, "num_tokens": 564464713.0, "step": 14799 }, { "epoch": 1.8827121231395498, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.119315147399902, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8590965867042542, "num_tokens": 564500667.0, "step": 14800 }, { "epoch": 1.8828393334181401, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.126614570617676, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8788903951644897, "num_tokens": 564541460.0, "step": 14801 }, { "epoch": 1.8829665436967307, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.15009880065918, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8620070815086365, "num_tokens": 564580983.0, "step": 14802 }, { "epoch": 1.8830937539753212, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.1045560836792, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8786643147468567, "num_tokens": 564615967.0, "step": 14803 }, { "epoch": 1.8832209642539117, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.17061710357666, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.861648678779602, "num_tokens": 564653255.0, "step": 14804 }, { "epoch": 1.8833481745325023, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.153129577636719, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8682379722595215, "num_tokens": 564688420.0, "step": 14805 }, { "epoch": 1.8834753848110928, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.142058372497559, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8642109632492065, "num_tokens": 564733123.0, "step": 14806 }, { "epoch": 1.883602595089683, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.070809364318848, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8772785663604736, "num_tokens": 564767712.0, "step": 14807 }, { "epoch": 1.8837298053682736, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.128393173217773, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8669579029083252, "num_tokens": 564810663.0, "step": 14808 }, { "epoch": 1.8838570156468641, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.088376998901367, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8681464791297913, "num_tokens": 564848689.0, "step": 14809 }, { "epoch": 1.8839842259254547, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.133508682250977, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8582603931427002, "num_tokens": 564887426.0, "step": 14810 }, { "epoch": 1.8841114362040452, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.034607887268066, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8741955757141113, "num_tokens": 564924852.0, "step": 14811 }, { "epoch": 1.8842386464826357, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.088118553161621, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8668029308319092, "num_tokens": 564964201.0, "step": 14812 }, { "epoch": 1.8843658567612263, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.116307258605957, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8807485103607178, "num_tokens": 565004939.0, "step": 14813 }, { "epoch": 1.8844930670398168, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.078591346740723, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8746935129165649, "num_tokens": 565044067.0, "step": 14814 }, { "epoch": 1.8846202773184073, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.128500938415527, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8714507818222046, "num_tokens": 565082114.0, "step": 14815 }, { "epoch": 1.8847474875969978, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.125295639038086, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8694537878036499, "num_tokens": 565118350.0, "step": 14816 }, { "epoch": 1.8848746978755884, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.158119201660156, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.862175464630127, "num_tokens": 565153313.0, "step": 14817 }, { "epoch": 1.885001908154179, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.089468002319336, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8572275638580322, "num_tokens": 565191332.0, "step": 14818 }, { "epoch": 1.8851291184327694, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.180907249450684, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8658618330955505, "num_tokens": 565229937.0, "step": 14819 }, { "epoch": 1.88525632871136, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.017643928527832, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8638592958450317, "num_tokens": 565266271.0, "step": 14820 }, { "epoch": 1.8853835389899505, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.110587120056152, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8670032620429993, "num_tokens": 565300593.0, "step": 14821 }, { "epoch": 1.885510749268541, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.111621856689453, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.869278609752655, "num_tokens": 565341801.0, "step": 14822 }, { "epoch": 1.8856379595471315, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.1635160446167, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8876926898956299, "num_tokens": 565374380.0, "step": 14823 }, { "epoch": 1.885765169825722, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.177592277526855, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8621665835380554, "num_tokens": 565413096.0, "step": 14824 }, { "epoch": 1.8858923801043126, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.15428352355957, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8794469833374023, "num_tokens": 565453292.0, "step": 14825 }, { "epoch": 1.886019590382903, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.15703010559082, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.865106463432312, "num_tokens": 565489311.0, "step": 14826 }, { "epoch": 1.8861468006614934, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.2103271484375, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8540768623352051, "num_tokens": 565526293.0, "step": 14827 }, { "epoch": 1.886274010940084, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.082527160644531, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8631357550621033, "num_tokens": 565569215.0, "step": 14828 }, { "epoch": 1.8864012212186745, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.114546775817871, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8806042075157166, "num_tokens": 565610755.0, "step": 14829 }, { "epoch": 1.886528431497265, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.148808479309082, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8752409815788269, "num_tokens": 565641964.0, "step": 14830 }, { "epoch": 1.8866556417758555, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.181571006774902, "learning_rate": 1e-06, "loss": 0.5096, "mean_token_accuracy": 0.8366159200668335, "num_tokens": 565682852.0, "step": 14831 }, { "epoch": 1.8867828520544458, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.08188247680664, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8857622146606445, "num_tokens": 565721352.0, "step": 14832 }, { "epoch": 1.8869100623330364, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.132408142089844, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8740242123603821, "num_tokens": 565764045.0, "step": 14833 }, { "epoch": 1.887037272611627, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.175293922424316, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8780503273010254, "num_tokens": 565797617.0, "step": 14834 }, { "epoch": 1.8871644828902174, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.166687965393066, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8689413070678711, "num_tokens": 565839254.0, "step": 14835 }, { "epoch": 1.887291693168808, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.132046699523926, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8701326847076416, "num_tokens": 565874742.0, "step": 14836 }, { "epoch": 1.8874189034473985, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.228076934814453, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8705224990844727, "num_tokens": 565911043.0, "step": 14837 }, { "epoch": 1.887546113725989, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.124812126159668, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8652734756469727, "num_tokens": 565950228.0, "step": 14838 }, { "epoch": 1.8876733240045795, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.146809577941895, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8813297748565674, "num_tokens": 565982891.0, "step": 14839 }, { "epoch": 1.88780053428317, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.187871932983398, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8681066036224365, "num_tokens": 566022601.0, "step": 14840 }, { "epoch": 1.8879277445617606, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.113670349121094, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8714698553085327, "num_tokens": 566054928.0, "step": 14841 }, { "epoch": 1.8880549548403511, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.17104721069336, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8631187081336975, "num_tokens": 566094900.0, "step": 14842 }, { "epoch": 1.8881821651189417, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.152410507202148, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8675727248191833, "num_tokens": 566125230.0, "step": 14843 }, { "epoch": 1.8883093753975322, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.160579681396484, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8847245573997498, "num_tokens": 566165320.0, "step": 14844 }, { "epoch": 1.8884365856761227, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.102276802062988, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8601680994033813, "num_tokens": 566202769.0, "step": 14845 }, { "epoch": 1.8885637959547132, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.166496276855469, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8719760179519653, "num_tokens": 566245874.0, "step": 14846 }, { "epoch": 1.8886910062333038, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.086187362670898, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8521665334701538, "num_tokens": 566284855.0, "step": 14847 }, { "epoch": 1.8888182165118943, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.09257698059082, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8613376021385193, "num_tokens": 566323422.0, "step": 14848 }, { "epoch": 1.8889454267904848, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.098625183105469, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8769819736480713, "num_tokens": 566364357.0, "step": 14849 }, { "epoch": 1.8890726370690751, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.135164260864258, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8666121959686279, "num_tokens": 566404498.0, "step": 14850 }, { "epoch": 1.8891998473476657, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.173460960388184, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8662450313568115, "num_tokens": 566442960.0, "step": 14851 }, { "epoch": 1.8893270576262562, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.150378227233887, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8787120580673218, "num_tokens": 566475087.0, "step": 14852 }, { "epoch": 1.8894542679048467, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.121230125427246, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.860785722732544, "num_tokens": 566514337.0, "step": 14853 }, { "epoch": 1.8895814781834372, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.193106651306152, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8599786758422852, "num_tokens": 566553204.0, "step": 14854 }, { "epoch": 1.8897086884620278, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.146034240722656, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8676411509513855, "num_tokens": 566589697.0, "step": 14855 }, { "epoch": 1.889835898740618, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.170637130737305, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8561515212059021, "num_tokens": 566629680.0, "step": 14856 }, { "epoch": 1.8899631090192086, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.099861145019531, "learning_rate": 1e-06, "loss": 0.3327, "mean_token_accuracy": 0.8929609060287476, "num_tokens": 566670931.0, "step": 14857 }, { "epoch": 1.8900903192977991, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.228715896606445, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.864094614982605, "num_tokens": 566706673.0, "step": 14858 }, { "epoch": 1.8902175295763897, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.119978904724121, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8648298978805542, "num_tokens": 566749503.0, "step": 14859 }, { "epoch": 1.8903447398549802, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.184162139892578, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8839086890220642, "num_tokens": 566790700.0, "step": 14860 }, { "epoch": 1.8904719501335707, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.174891471862793, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8693450689315796, "num_tokens": 566832995.0, "step": 14861 }, { "epoch": 1.8905991604121613, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.175910949707031, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8578663468360901, "num_tokens": 566868190.0, "step": 14862 }, { "epoch": 1.8907263706907518, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.17294979095459, "learning_rate": 1e-06, "loss": 0.4902, "mean_token_accuracy": 0.8417074680328369, "num_tokens": 566906637.0, "step": 14863 }, { "epoch": 1.8908535809693423, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.087005615234375, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8698970079421997, "num_tokens": 566946006.0, "step": 14864 }, { "epoch": 1.8909807912479328, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.189241409301758, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8573697805404663, "num_tokens": 566987726.0, "step": 14865 }, { "epoch": 1.8911080015265234, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.056323051452637, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8842914700508118, "num_tokens": 567027432.0, "step": 14866 }, { "epoch": 1.891235211805114, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.193079948425293, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8756752610206604, "num_tokens": 567060607.0, "step": 14867 }, { "epoch": 1.8913624220837044, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.15243911743164, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8631383776664734, "num_tokens": 567100691.0, "step": 14868 }, { "epoch": 1.891489632362295, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.125991821289062, "learning_rate": 1e-06, "loss": 0.3464, "mean_token_accuracy": 0.8883269429206848, "num_tokens": 567139460.0, "step": 14869 }, { "epoch": 1.8916168426408855, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.101136207580566, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8576966524124146, "num_tokens": 567185384.0, "step": 14870 }, { "epoch": 1.891744052919476, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.109516143798828, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8693674206733704, "num_tokens": 567219898.0, "step": 14871 }, { "epoch": 1.8918712631980665, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.056134223937988, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8772810697555542, "num_tokens": 567260322.0, "step": 14872 }, { "epoch": 1.891998473476657, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.080222129821777, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8649897575378418, "num_tokens": 567293739.0, "step": 14873 }, { "epoch": 1.8921256837552476, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.149530410766602, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8712759017944336, "num_tokens": 567327282.0, "step": 14874 }, { "epoch": 1.892252894033838, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.156044960021973, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.863014817237854, "num_tokens": 567366402.0, "step": 14875 }, { "epoch": 1.8923801043124284, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.154600143432617, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8616082072257996, "num_tokens": 567409924.0, "step": 14876 }, { "epoch": 1.892507314591019, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.111513137817383, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8721504211425781, "num_tokens": 567450355.0, "step": 14877 }, { "epoch": 1.8926345248696095, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.16285514831543, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8831289410591125, "num_tokens": 567487701.0, "step": 14878 }, { "epoch": 1.8927617351482, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.128694534301758, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8684704303741455, "num_tokens": 567527619.0, "step": 14879 }, { "epoch": 1.8928889454267905, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.196590423583984, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8685498833656311, "num_tokens": 567567016.0, "step": 14880 }, { "epoch": 1.8930161557053808, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.135666847229004, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8569895625114441, "num_tokens": 567605465.0, "step": 14881 }, { "epoch": 1.8931433659839714, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.155538558959961, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8719886541366577, "num_tokens": 567646680.0, "step": 14882 }, { "epoch": 1.893270576262562, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.137662887573242, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8703270554542542, "num_tokens": 567685685.0, "step": 14883 }, { "epoch": 1.8933977865411524, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.212088584899902, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8586657047271729, "num_tokens": 567717222.0, "step": 14884 }, { "epoch": 1.893524996819743, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.168310165405273, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8600281476974487, "num_tokens": 567761050.0, "step": 14885 }, { "epoch": 1.8936522070983335, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.094880104064941, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.867348313331604, "num_tokens": 567801423.0, "step": 14886 }, { "epoch": 1.893779417376924, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.146812438964844, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.860165536403656, "num_tokens": 567844614.0, "step": 14887 }, { "epoch": 1.8939066276555145, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.132830619812012, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8724186420440674, "num_tokens": 567878803.0, "step": 14888 }, { "epoch": 1.894033837934105, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.166606903076172, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.881147027015686, "num_tokens": 567909419.0, "step": 14889 }, { "epoch": 1.8941610482126956, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.119251251220703, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8702470660209656, "num_tokens": 567949221.0, "step": 14890 }, { "epoch": 1.8942882584912861, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.160232543945312, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.885200023651123, "num_tokens": 567985101.0, "step": 14891 }, { "epoch": 1.8944154687698767, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.145798683166504, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8522434234619141, "num_tokens": 568030542.0, "step": 14892 }, { "epoch": 1.8945426790484672, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.10550594329834, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.872099757194519, "num_tokens": 568070723.0, "step": 14893 }, { "epoch": 1.8946698893270577, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.206011772155762, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8629995584487915, "num_tokens": 568113444.0, "step": 14894 }, { "epoch": 1.8947970996056482, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.104743003845215, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8731299042701721, "num_tokens": 568152646.0, "step": 14895 }, { "epoch": 1.8949243098842388, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.180221557617188, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8694727420806885, "num_tokens": 568187046.0, "step": 14896 }, { "epoch": 1.8950515201628293, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.160749435424805, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8649730682373047, "num_tokens": 568225052.0, "step": 14897 }, { "epoch": 1.8951787304414198, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.15557861328125, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8614914417266846, "num_tokens": 568258299.0, "step": 14898 }, { "epoch": 1.8953059407200101, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.253040313720703, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8756663799285889, "num_tokens": 568297137.0, "step": 14899 }, { "epoch": 1.8954331509986007, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.11241626739502, "learning_rate": 1e-06, "loss": 0.3382, "mean_token_accuracy": 0.8907821178436279, "num_tokens": 568331102.0, "step": 14900 }, { "epoch": 1.8955603612771912, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.174844741821289, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8670560121536255, "num_tokens": 568373926.0, "step": 14901 }, { "epoch": 1.8956875715557817, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.174057006835938, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8791210055351257, "num_tokens": 568406470.0, "step": 14902 }, { "epoch": 1.8958147818343722, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.179389953613281, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8767113089561462, "num_tokens": 568443452.0, "step": 14903 }, { "epoch": 1.8959419921129628, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.241189002990723, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8649902939796448, "num_tokens": 568480085.0, "step": 14904 }, { "epoch": 1.896069202391553, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.18616771697998, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8673486709594727, "num_tokens": 568519191.0, "step": 14905 }, { "epoch": 1.8961964126701436, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.177240371704102, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8753751516342163, "num_tokens": 568554224.0, "step": 14906 }, { "epoch": 1.8963236229487341, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.231010437011719, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.886534571647644, "num_tokens": 568586102.0, "step": 14907 }, { "epoch": 1.8964508332273247, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.1983003616333, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8641725778579712, "num_tokens": 568625224.0, "step": 14908 }, { "epoch": 1.8965780435059152, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.256278038024902, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8479148149490356, "num_tokens": 568665588.0, "step": 14909 }, { "epoch": 1.8967052537845057, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.182923316955566, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8852931261062622, "num_tokens": 568698408.0, "step": 14910 }, { "epoch": 1.8968324640630962, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.194665908813477, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.857365071773529, "num_tokens": 568731982.0, "step": 14911 }, { "epoch": 1.8969596743416868, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.186986923217773, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8691925406455994, "num_tokens": 568769388.0, "step": 14912 }, { "epoch": 1.8970868846202773, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.1843843460083, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8767863512039185, "num_tokens": 568810329.0, "step": 14913 }, { "epoch": 1.8972140948988678, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.15302848815918, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8746013641357422, "num_tokens": 568852241.0, "step": 14914 }, { "epoch": 1.8973413051774584, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.201635360717773, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8645838499069214, "num_tokens": 568889660.0, "step": 14915 }, { "epoch": 1.8974685154560489, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.181634902954102, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8674968481063843, "num_tokens": 568934489.0, "step": 14916 }, { "epoch": 1.8975957257346394, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.171111106872559, "learning_rate": 1e-06, "loss": 0.3353, "mean_token_accuracy": 0.891572892665863, "num_tokens": 568971904.0, "step": 14917 }, { "epoch": 1.89772293601323, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.247395515441895, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8515865802764893, "num_tokens": 569011863.0, "step": 14918 }, { "epoch": 1.8978501462918205, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.155082702636719, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8912331461906433, "num_tokens": 569050295.0, "step": 14919 }, { "epoch": 1.897977356570411, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.171661376953125, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.863179087638855, "num_tokens": 569087013.0, "step": 14920 }, { "epoch": 1.8981045668490015, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.251619338989258, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8628759384155273, "num_tokens": 569125226.0, "step": 14921 }, { "epoch": 1.898231777127592, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.179903030395508, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8655847311019897, "num_tokens": 569161910.0, "step": 14922 }, { "epoch": 1.8983589874061826, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.191969871520996, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8727734088897705, "num_tokens": 569196519.0, "step": 14923 }, { "epoch": 1.898486197684773, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.198992729187012, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8650932312011719, "num_tokens": 569234404.0, "step": 14924 }, { "epoch": 1.8986134079633634, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.178240776062012, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8617866039276123, "num_tokens": 569272728.0, "step": 14925 }, { "epoch": 1.898740618241954, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.166670799255371, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.861457347869873, "num_tokens": 569315434.0, "step": 14926 }, { "epoch": 1.8988678285205445, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.099563598632812, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8530361652374268, "num_tokens": 569357473.0, "step": 14927 }, { "epoch": 1.898995038799135, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.13638687133789, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8580967783927917, "num_tokens": 569398747.0, "step": 14928 }, { "epoch": 1.8991222490777255, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.211165428161621, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8770777583122253, "num_tokens": 569434855.0, "step": 14929 }, { "epoch": 1.8992494593563158, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.079805374145508, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8691455125808716, "num_tokens": 569475236.0, "step": 14930 }, { "epoch": 1.8993766696349064, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.15257740020752, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8678644895553589, "num_tokens": 569513599.0, "step": 14931 }, { "epoch": 1.899503879913497, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.191485404968262, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8758432269096375, "num_tokens": 569559059.0, "step": 14932 }, { "epoch": 1.8996310901920874, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.105857849121094, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.886516809463501, "num_tokens": 569595240.0, "step": 14933 }, { "epoch": 1.899758300470678, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.152053833007812, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.859606146812439, "num_tokens": 569629336.0, "step": 14934 }, { "epoch": 1.8998855107492685, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.149860382080078, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8806250095367432, "num_tokens": 569663613.0, "step": 14935 }, { "epoch": 1.900012721027859, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.17037582397461, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8537330627441406, "num_tokens": 569702884.0, "step": 14936 }, { "epoch": 1.9001399313064495, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.136377334594727, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8837462663650513, "num_tokens": 569744618.0, "step": 14937 }, { "epoch": 1.90026714158504, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.14700698852539, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8785902261734009, "num_tokens": 569777083.0, "step": 14938 }, { "epoch": 1.9003943518636306, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.192150115966797, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8767867684364319, "num_tokens": 569809652.0, "step": 14939 }, { "epoch": 1.9005215621422211, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.21095085144043, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8738805055618286, "num_tokens": 569846465.0, "step": 14940 }, { "epoch": 1.9006487724208116, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.150126457214355, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8625586628913879, "num_tokens": 569883655.0, "step": 14941 }, { "epoch": 1.9007759826994022, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.191055297851562, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8685401678085327, "num_tokens": 569923358.0, "step": 14942 }, { "epoch": 1.9009031929779927, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.175785064697266, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8664160370826721, "num_tokens": 569961393.0, "step": 14943 }, { "epoch": 1.9010304032565832, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.173232078552246, "learning_rate": 1e-06, "loss": 0.4945, "mean_token_accuracy": 0.8424324989318848, "num_tokens": 570002092.0, "step": 14944 }, { "epoch": 1.9011576135351738, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.191803932189941, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.855785071849823, "num_tokens": 570037699.0, "step": 14945 }, { "epoch": 1.9012848238137643, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.116806983947754, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8472667932510376, "num_tokens": 570077693.0, "step": 14946 }, { "epoch": 1.9014120340923548, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.185145378112793, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8889293670654297, "num_tokens": 570115354.0, "step": 14947 }, { "epoch": 1.9015392443709451, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.162079811096191, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8775725364685059, "num_tokens": 570154387.0, "step": 14948 }, { "epoch": 1.9016664546495357, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.179110527038574, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.86419677734375, "num_tokens": 570189743.0, "step": 14949 }, { "epoch": 1.9017936649281262, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.19273853302002, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8817952871322632, "num_tokens": 570224009.0, "step": 14950 }, { "epoch": 1.9019208752067167, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.168163299560547, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.87635737657547, "num_tokens": 570266703.0, "step": 14951 }, { "epoch": 1.9020480854853072, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.211740493774414, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8689342737197876, "num_tokens": 570308241.0, "step": 14952 }, { "epoch": 1.9021752957638978, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.189571380615234, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8695571422576904, "num_tokens": 570346931.0, "step": 14953 }, { "epoch": 1.902302506042488, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.160404205322266, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8833321928977966, "num_tokens": 570380373.0, "step": 14954 }, { "epoch": 1.9024297163210786, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.189848899841309, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8807747960090637, "num_tokens": 570420106.0, "step": 14955 }, { "epoch": 1.9025569265996691, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.20818042755127, "learning_rate": 1e-06, "loss": 0.3477, "mean_token_accuracy": 0.8912428617477417, "num_tokens": 570464534.0, "step": 14956 }, { "epoch": 1.9026841368782597, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.19741153717041, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8593641519546509, "num_tokens": 570505734.0, "step": 14957 }, { "epoch": 1.9028113471568502, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.139119148254395, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8792396783828735, "num_tokens": 570545913.0, "step": 14958 }, { "epoch": 1.9029385574354407, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.23643970489502, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8764584064483643, "num_tokens": 570584397.0, "step": 14959 }, { "epoch": 1.9030657677140312, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.117679595947266, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8755239248275757, "num_tokens": 570622285.0, "step": 14960 }, { "epoch": 1.9031929779926218, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.172162055969238, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8549728989601135, "num_tokens": 570656331.0, "step": 14961 }, { "epoch": 1.9033201882712123, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.203146934509277, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8737064599990845, "num_tokens": 570699944.0, "step": 14962 }, { "epoch": 1.9034473985498028, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.138230323791504, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8640279173851013, "num_tokens": 570737649.0, "step": 14963 }, { "epoch": 1.9035746088283934, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.18537712097168, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8668018579483032, "num_tokens": 570773193.0, "step": 14964 }, { "epoch": 1.9037018191069839, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.14886474609375, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8630315065383911, "num_tokens": 570809117.0, "step": 14965 }, { "epoch": 1.9038290293855744, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.136698722839355, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8809646964073181, "num_tokens": 570844268.0, "step": 14966 }, { "epoch": 1.903956239664165, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.14028263092041, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8612174391746521, "num_tokens": 570890166.0, "step": 14967 }, { "epoch": 1.9040834499427555, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.150593757629395, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8650888800621033, "num_tokens": 570921419.0, "step": 14968 }, { "epoch": 1.904210660221346, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.180797576904297, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8600763082504272, "num_tokens": 570966444.0, "step": 14969 }, { "epoch": 1.9043378704999365, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.119407653808594, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8800990581512451, "num_tokens": 571011656.0, "step": 14970 }, { "epoch": 1.904465080778527, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.231464385986328, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8735857009887695, "num_tokens": 571047828.0, "step": 14971 }, { "epoch": 1.9045922910571176, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.163933753967285, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.867892861366272, "num_tokens": 571084541.0, "step": 14972 }, { "epoch": 1.9047195013357079, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.209293365478516, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8642269372940063, "num_tokens": 571120978.0, "step": 14973 }, { "epoch": 1.9048467116142984, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.202020645141602, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8830519914627075, "num_tokens": 571160181.0, "step": 14974 }, { "epoch": 1.904973921892889, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.174100875854492, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8749585747718811, "num_tokens": 571194648.0, "step": 14975 }, { "epoch": 1.9051011321714795, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.193355560302734, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8648414015769958, "num_tokens": 571238449.0, "step": 14976 }, { "epoch": 1.90522834245007, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.221344947814941, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8714044094085693, "num_tokens": 571277266.0, "step": 14977 }, { "epoch": 1.9053555527286605, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.179466247558594, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8808643221855164, "num_tokens": 571316891.0, "step": 14978 }, { "epoch": 1.9054827630072508, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.17013931274414, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8713784217834473, "num_tokens": 571353097.0, "step": 14979 }, { "epoch": 1.9056099732858414, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.232098579406738, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8840703964233398, "num_tokens": 571392217.0, "step": 14980 }, { "epoch": 1.905737183564432, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.176861763000488, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8782454133033752, "num_tokens": 571430440.0, "step": 14981 }, { "epoch": 1.9058643938430224, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.219066619873047, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8658624291419983, "num_tokens": 571472002.0, "step": 14982 }, { "epoch": 1.905991604121613, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.176704406738281, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.8510255217552185, "num_tokens": 571509368.0, "step": 14983 }, { "epoch": 1.9061188144002035, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.173421859741211, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8553777933120728, "num_tokens": 571543898.0, "step": 14984 }, { "epoch": 1.906246024678794, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.157454490661621, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8644213676452637, "num_tokens": 571584395.0, "step": 14985 }, { "epoch": 1.9063732349573845, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.233389854431152, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8692963123321533, "num_tokens": 571624792.0, "step": 14986 }, { "epoch": 1.906500445235975, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.149908065795898, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8717830181121826, "num_tokens": 571667197.0, "step": 14987 }, { "epoch": 1.9066276555145656, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.166738510131836, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8642066717147827, "num_tokens": 571703648.0, "step": 14988 }, { "epoch": 1.9067548657931561, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.199111938476562, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8617483377456665, "num_tokens": 571738980.0, "step": 14989 }, { "epoch": 1.9068820760717466, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.175145149230957, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8722743988037109, "num_tokens": 571774540.0, "step": 14990 }, { "epoch": 1.9070092863503372, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.227928161621094, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.87459397315979, "num_tokens": 571805286.0, "step": 14991 }, { "epoch": 1.9071364966289277, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.182108879089355, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8640210032463074, "num_tokens": 571847936.0, "step": 14992 }, { "epoch": 1.9072637069075182, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.17635726928711, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8653036952018738, "num_tokens": 571885454.0, "step": 14993 }, { "epoch": 1.9073909171861088, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.185830116271973, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8685281276702881, "num_tokens": 571924870.0, "step": 14994 }, { "epoch": 1.9075181274646993, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.202685356140137, "learning_rate": 1e-06, "loss": 0.4811, "mean_token_accuracy": 0.850120484828949, "num_tokens": 571957870.0, "step": 14995 }, { "epoch": 1.9076453377432898, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.199822425842285, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8618587255477905, "num_tokens": 571993645.0, "step": 14996 }, { "epoch": 1.9077725480218801, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.22339916229248, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8782776594161987, "num_tokens": 572028594.0, "step": 14997 }, { "epoch": 1.9078997583004706, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.146333694458008, "learning_rate": 1e-06, "loss": 0.489, "mean_token_accuracy": 0.8437467217445374, "num_tokens": 572066971.0, "step": 14998 }, { "epoch": 1.9080269685790612, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.232560157775879, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8729601502418518, "num_tokens": 572107609.0, "step": 14999 }, { "epoch": 1.9081541788576517, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.181011199951172, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8780558109283447, "num_tokens": 572146940.0, "step": 15000 }, { "epoch": 1.9082813891362422, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.221710205078125, "learning_rate": 1e-06, "loss": 0.3313, "mean_token_accuracy": 0.8943123817443848, "num_tokens": 572184010.0, "step": 15001 }, { "epoch": 1.9084085994148328, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.208504676818848, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8538810610771179, "num_tokens": 572223184.0, "step": 15002 }, { "epoch": 1.908535809693423, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.206169128417969, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8695266246795654, "num_tokens": 572261770.0, "step": 15003 }, { "epoch": 1.9086630199720136, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.149930953979492, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8727591037750244, "num_tokens": 572303229.0, "step": 15004 }, { "epoch": 1.9087902302506041, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.248893737792969, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8572954535484314, "num_tokens": 572337839.0, "step": 15005 }, { "epoch": 1.9089174405291947, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.187182426452637, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8689707517623901, "num_tokens": 572373604.0, "step": 15006 }, { "epoch": 1.9090446508077852, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.18803882598877, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8770699501037598, "num_tokens": 572410425.0, "step": 15007 }, { "epoch": 1.9091718610863757, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.208691596984863, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8781823515892029, "num_tokens": 572443998.0, "step": 15008 }, { "epoch": 1.9092990713649662, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.208664894104004, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8574000000953674, "num_tokens": 572485337.0, "step": 15009 }, { "epoch": 1.9094262816435568, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.227080345153809, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8591718077659607, "num_tokens": 572528273.0, "step": 15010 }, { "epoch": 1.9095534919221473, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.194649696350098, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.872403085231781, "num_tokens": 572569467.0, "step": 15011 }, { "epoch": 1.9096807022007378, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.18702507019043, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.877174437046051, "num_tokens": 572609626.0, "step": 15012 }, { "epoch": 1.9098079124793284, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.2127685546875, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8535553216934204, "num_tokens": 572652907.0, "step": 15013 }, { "epoch": 1.9099351227579189, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.18593692779541, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8747460842132568, "num_tokens": 572687460.0, "step": 15014 }, { "epoch": 1.9100623330365094, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.308286666870117, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8719955682754517, "num_tokens": 572723937.0, "step": 15015 }, { "epoch": 1.9101895433151, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.135872840881348, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8656728863716125, "num_tokens": 572764929.0, "step": 15016 }, { "epoch": 1.9103167535936905, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.265382766723633, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.866724967956543, "num_tokens": 572802545.0, "step": 15017 }, { "epoch": 1.910443963872281, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.1371488571167, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8636641502380371, "num_tokens": 572850008.0, "step": 15018 }, { "epoch": 1.9105711741508715, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.204773902893066, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8775612115859985, "num_tokens": 572888968.0, "step": 15019 }, { "epoch": 1.910698384429462, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.231144905090332, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8683738708496094, "num_tokens": 572926314.0, "step": 15020 }, { "epoch": 1.9108255947080524, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.209869384765625, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8730383515357971, "num_tokens": 572961789.0, "step": 15021 }, { "epoch": 1.9109528049866429, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.177201271057129, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8710613250732422, "num_tokens": 572993680.0, "step": 15022 }, { "epoch": 1.9110800152652334, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.194942474365234, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8719145059585571, "num_tokens": 573024003.0, "step": 15023 }, { "epoch": 1.911207225543824, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.15008544921875, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8772474527359009, "num_tokens": 573055138.0, "step": 15024 }, { "epoch": 1.9113344358224145, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.170358657836914, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.874610424041748, "num_tokens": 573092373.0, "step": 15025 }, { "epoch": 1.911461646101005, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.120817184448242, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8643707036972046, "num_tokens": 573131486.0, "step": 15026 }, { "epoch": 1.9115888563795955, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.203744888305664, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8794952630996704, "num_tokens": 573170114.0, "step": 15027 }, { "epoch": 1.9117160666581858, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.168962478637695, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8731045722961426, "num_tokens": 573200906.0, "step": 15028 }, { "epoch": 1.9118432769367764, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.132312774658203, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8564116954803467, "num_tokens": 573243731.0, "step": 15029 }, { "epoch": 1.9119704872153669, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.174534797668457, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8759259581565857, "num_tokens": 573283335.0, "step": 15030 }, { "epoch": 1.9120976974939574, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.214193344116211, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8573272824287415, "num_tokens": 573326628.0, "step": 15031 }, { "epoch": 1.912224907772548, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.13199234008789, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8755799531936646, "num_tokens": 573369025.0, "step": 15032 }, { "epoch": 1.9123521180511385, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.2257080078125, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8631539940834045, "num_tokens": 573411923.0, "step": 15033 }, { "epoch": 1.912479328329729, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.157654762268066, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8652338981628418, "num_tokens": 573450455.0, "step": 15034 }, { "epoch": 1.9126065386083195, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.161051750183105, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8601557016372681, "num_tokens": 573492886.0, "step": 15035 }, { "epoch": 1.91273374888691, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.193062782287598, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8739334940910339, "num_tokens": 573532073.0, "step": 15036 }, { "epoch": 1.9128609591655006, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.133852005004883, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8855633735656738, "num_tokens": 573573632.0, "step": 15037 }, { "epoch": 1.9129881694440911, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.22225284576416, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8794400691986084, "num_tokens": 573610194.0, "step": 15038 }, { "epoch": 1.9131153797226816, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.227415084838867, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8616054058074951, "num_tokens": 573650237.0, "step": 15039 }, { "epoch": 1.9132425900012722, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.144038200378418, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8737286329269409, "num_tokens": 573686859.0, "step": 15040 }, { "epoch": 1.9133698002798627, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.168781280517578, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8697057962417603, "num_tokens": 573726770.0, "step": 15041 }, { "epoch": 1.9134970105584532, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.226035118103027, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8639063835144043, "num_tokens": 573768668.0, "step": 15042 }, { "epoch": 1.9136242208370438, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.1436185836792, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8605614304542542, "num_tokens": 573807880.0, "step": 15043 }, { "epoch": 1.9137514311156343, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.213223457336426, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8675969839096069, "num_tokens": 573847018.0, "step": 15044 }, { "epoch": 1.9138786413942248, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.196616172790527, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8847070932388306, "num_tokens": 573881193.0, "step": 15045 }, { "epoch": 1.9140058516728151, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.151426315307617, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8782673478126526, "num_tokens": 573918457.0, "step": 15046 }, { "epoch": 1.9141330619514056, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.271936416625977, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8886305093765259, "num_tokens": 573958818.0, "step": 15047 }, { "epoch": 1.9142602722299962, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.101428985595703, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8639517426490784, "num_tokens": 573996735.0, "step": 15048 }, { "epoch": 1.9143874825085867, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.20241641998291, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8739275932312012, "num_tokens": 574035005.0, "step": 15049 }, { "epoch": 1.9145146927871772, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.1909761428833, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8686975240707397, "num_tokens": 574077817.0, "step": 15050 }, { "epoch": 1.9146419030657678, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.277969360351562, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8774909377098083, "num_tokens": 574113083.0, "step": 15051 }, { "epoch": 1.914769113344358, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.145627975463867, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8739439249038696, "num_tokens": 574147324.0, "step": 15052 }, { "epoch": 1.9148963236229486, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 13.157342910766602, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8843443989753723, "num_tokens": 574186108.0, "step": 15053 }, { "epoch": 1.9150235339015391, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.2653226852417, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8831269145011902, "num_tokens": 574218076.0, "step": 15054 }, { "epoch": 1.9151507441801296, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.078611373901367, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8796504735946655, "num_tokens": 574256416.0, "step": 15055 }, { "epoch": 1.9152779544587202, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.195213317871094, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8839238882064819, "num_tokens": 574295262.0, "step": 15056 }, { "epoch": 1.9154051647373107, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.26521110534668, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.860002338886261, "num_tokens": 574329836.0, "step": 15057 }, { "epoch": 1.9155323750159012, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.125746726989746, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8729065656661987, "num_tokens": 574363189.0, "step": 15058 }, { "epoch": 1.9156595852944918, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.251731872558594, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8781877160072327, "num_tokens": 574396374.0, "step": 15059 }, { "epoch": 1.9157867955730823, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.11890697479248, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8727429509162903, "num_tokens": 574435488.0, "step": 15060 }, { "epoch": 1.9159140058516728, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.201855659484863, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8706562519073486, "num_tokens": 574475261.0, "step": 15061 }, { "epoch": 1.9160412161302633, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.191195487976074, "learning_rate": 1e-06, "loss": 0.3433, "mean_token_accuracy": 0.8890471458435059, "num_tokens": 574513531.0, "step": 15062 }, { "epoch": 1.9161684264088539, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.179814338684082, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.856022834777832, "num_tokens": 574551736.0, "step": 15063 }, { "epoch": 1.9162956366874444, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.195109367370605, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8645114898681641, "num_tokens": 574589978.0, "step": 15064 }, { "epoch": 1.916422846966035, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.153152465820312, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8496419787406921, "num_tokens": 574628845.0, "step": 15065 }, { "epoch": 1.9165500572446255, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.204996109008789, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8663874864578247, "num_tokens": 574660700.0, "step": 15066 }, { "epoch": 1.916677267523216, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.204231262207031, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8614965677261353, "num_tokens": 574692195.0, "step": 15067 }, { "epoch": 1.9168044778018065, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.144134521484375, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8602631092071533, "num_tokens": 574732462.0, "step": 15068 }, { "epoch": 1.916931688080397, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.176599502563477, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8769224882125854, "num_tokens": 574769353.0, "step": 15069 }, { "epoch": 1.9170588983589874, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.215874671936035, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8706403374671936, "num_tokens": 574804958.0, "step": 15070 }, { "epoch": 1.9171861086375779, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.222904205322266, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8696688413619995, "num_tokens": 574843966.0, "step": 15071 }, { "epoch": 1.9173133189161684, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.172104835510254, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8743370771408081, "num_tokens": 574882916.0, "step": 15072 }, { "epoch": 1.917440529194759, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.193731307983398, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8789495825767517, "num_tokens": 574922154.0, "step": 15073 }, { "epoch": 1.9175677394733495, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.197343826293945, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8668224811553955, "num_tokens": 574956894.0, "step": 15074 }, { "epoch": 1.91769494975194, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.162376403808594, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.884800910949707, "num_tokens": 574998343.0, "step": 15075 }, { "epoch": 1.9178221600305305, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.237191200256348, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8795017600059509, "num_tokens": 575036689.0, "step": 15076 }, { "epoch": 1.9179493703091208, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.165823936462402, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8651313781738281, "num_tokens": 575075510.0, "step": 15077 }, { "epoch": 1.9180765805877114, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.251057624816895, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8673883080482483, "num_tokens": 575115797.0, "step": 15078 }, { "epoch": 1.9182037908663019, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.198655128479004, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8568928241729736, "num_tokens": 575155240.0, "step": 15079 }, { "epoch": 1.9183310011448924, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.181206703186035, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8568680286407471, "num_tokens": 575191048.0, "step": 15080 }, { "epoch": 1.918458211423483, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.242548942565918, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8596125841140747, "num_tokens": 575227778.0, "step": 15081 }, { "epoch": 1.9185854217020735, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.151538848876953, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8717844486236572, "num_tokens": 575269774.0, "step": 15082 }, { "epoch": 1.918712631980664, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.141142845153809, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8617388606071472, "num_tokens": 575310710.0, "step": 15083 }, { "epoch": 1.9188398422592545, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.18027400970459, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8775539398193359, "num_tokens": 575344073.0, "step": 15084 }, { "epoch": 1.918967052537845, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.190367698669434, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.861073911190033, "num_tokens": 575386737.0, "step": 15085 }, { "epoch": 1.9190942628164356, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.191712379455566, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.875004768371582, "num_tokens": 575424287.0, "step": 15086 }, { "epoch": 1.919221473095026, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.207351684570312, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8716256618499756, "num_tokens": 575463859.0, "step": 15087 }, { "epoch": 1.9193486833736166, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.262809753417969, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8644837141036987, "num_tokens": 575500838.0, "step": 15088 }, { "epoch": 1.9194758936522072, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.21255111694336, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8625100255012512, "num_tokens": 575535164.0, "step": 15089 }, { "epoch": 1.9196031039307977, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.193944931030273, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8768008947372437, "num_tokens": 575569493.0, "step": 15090 }, { "epoch": 1.9197303142093882, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.204815864562988, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8548312783241272, "num_tokens": 575603396.0, "step": 15091 }, { "epoch": 1.9198575244879788, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.191169738769531, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8762811422348022, "num_tokens": 575638255.0, "step": 15092 }, { "epoch": 1.9199847347665693, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.17418384552002, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8602973222732544, "num_tokens": 575685281.0, "step": 15093 }, { "epoch": 1.9201119450451598, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.1810884475708, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8764129877090454, "num_tokens": 575725030.0, "step": 15094 }, { "epoch": 1.9202391553237501, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.159990310668945, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.865104079246521, "num_tokens": 575764237.0, "step": 15095 }, { "epoch": 1.9203663656023406, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.18486499786377, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8677088618278503, "num_tokens": 575796732.0, "step": 15096 }, { "epoch": 1.9204935758809312, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.15565013885498, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8743156790733337, "num_tokens": 575833133.0, "step": 15097 }, { "epoch": 1.9206207861595217, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.160592079162598, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8661175966262817, "num_tokens": 575870920.0, "step": 15098 }, { "epoch": 1.9207479964381122, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.185173988342285, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8728300333023071, "num_tokens": 575909132.0, "step": 15099 }, { "epoch": 1.9208752067167028, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.249772071838379, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8703569173812866, "num_tokens": 575947979.0, "step": 15100 }, { "epoch": 1.921002416995293, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.195194244384766, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8589837551116943, "num_tokens": 575988888.0, "step": 15101 }, { "epoch": 1.9211296272738836, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.165212631225586, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8722314834594727, "num_tokens": 576028614.0, "step": 15102 }, { "epoch": 1.9212568375524741, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.162511825561523, "learning_rate": 1e-06, "loss": 0.325, "mean_token_accuracy": 0.8974838256835938, "num_tokens": 576066648.0, "step": 15103 }, { "epoch": 1.9213840478310646, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.254637718200684, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8685779571533203, "num_tokens": 576104045.0, "step": 15104 }, { "epoch": 1.9215112581096552, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.152446746826172, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8632731437683105, "num_tokens": 576145419.0, "step": 15105 }, { "epoch": 1.9216384683882457, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.198235511779785, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8665574789047241, "num_tokens": 576191410.0, "step": 15106 }, { "epoch": 1.9217656786668362, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.226059913635254, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8649563789367676, "num_tokens": 576226383.0, "step": 15107 }, { "epoch": 1.9218928889454268, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.1770658493042, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8629779815673828, "num_tokens": 576259958.0, "step": 15108 }, { "epoch": 1.9220200992240173, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.214908599853516, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8724598288536072, "num_tokens": 576299467.0, "step": 15109 }, { "epoch": 1.9221473095026078, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.282439231872559, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.862892746925354, "num_tokens": 576334137.0, "step": 15110 }, { "epoch": 1.9222745197811983, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.159966468811035, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.866934061050415, "num_tokens": 576379582.0, "step": 15111 }, { "epoch": 1.9224017300597889, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.268061637878418, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8841965198516846, "num_tokens": 576414590.0, "step": 15112 }, { "epoch": 1.9225289403383794, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.18102741241455, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8614968061447144, "num_tokens": 576445023.0, "step": 15113 }, { "epoch": 1.92265615061697, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.258910179138184, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8661872744560242, "num_tokens": 576477861.0, "step": 15114 }, { "epoch": 1.9227833608955605, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.2302885055542, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8655993342399597, "num_tokens": 576513121.0, "step": 15115 }, { "epoch": 1.922910571174151, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.238740921020508, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8758541345596313, "num_tokens": 576551039.0, "step": 15116 }, { "epoch": 1.9230377814527415, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.225489616394043, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8624101877212524, "num_tokens": 576590651.0, "step": 15117 }, { "epoch": 1.923164991731332, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.18067741394043, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8609215617179871, "num_tokens": 576629119.0, "step": 15118 }, { "epoch": 1.9232922020099223, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.232522964477539, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8791928291320801, "num_tokens": 576664240.0, "step": 15119 }, { "epoch": 1.9234194122885129, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.248384475708008, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8626015186309814, "num_tokens": 576701946.0, "step": 15120 }, { "epoch": 1.9235466225671034, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.23249626159668, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8806842565536499, "num_tokens": 576733541.0, "step": 15121 }, { "epoch": 1.923673832845694, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.209371566772461, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8561950922012329, "num_tokens": 576770930.0, "step": 15122 }, { "epoch": 1.9238010431242845, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.243077278137207, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8692262768745422, "num_tokens": 576807455.0, "step": 15123 }, { "epoch": 1.923928253402875, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.203686714172363, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8774235248565674, "num_tokens": 576849577.0, "step": 15124 }, { "epoch": 1.9240554636814655, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.208030700683594, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8640576601028442, "num_tokens": 576883506.0, "step": 15125 }, { "epoch": 1.9241826739600558, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.226824760437012, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8527204394340515, "num_tokens": 576923892.0, "step": 15126 }, { "epoch": 1.9243098842386464, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.223203659057617, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8866905570030212, "num_tokens": 576970832.0, "step": 15127 }, { "epoch": 1.9244370945172369, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.173223495483398, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8518168926239014, "num_tokens": 577006990.0, "step": 15128 }, { "epoch": 1.9245643047958274, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.221369743347168, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8750717639923096, "num_tokens": 577044932.0, "step": 15129 }, { "epoch": 1.924691515074418, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.178715705871582, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8530492782592773, "num_tokens": 577076217.0, "step": 15130 }, { "epoch": 1.9248187253530085, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.112592697143555, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8854820728302002, "num_tokens": 577112696.0, "step": 15131 }, { "epoch": 1.924945935631599, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.191168785095215, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8691153526306152, "num_tokens": 577156397.0, "step": 15132 }, { "epoch": 1.9250731459101895, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.126194953918457, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8737329840660095, "num_tokens": 577194192.0, "step": 15133 }, { "epoch": 1.92520035618878, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.138810157775879, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8650369644165039, "num_tokens": 577232639.0, "step": 15134 }, { "epoch": 1.9253275664673706, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.130791664123535, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8787394762039185, "num_tokens": 577267349.0, "step": 15135 }, { "epoch": 1.925454776745961, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.175207138061523, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8649253249168396, "num_tokens": 577308329.0, "step": 15136 }, { "epoch": 1.9255819870245516, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.245928764343262, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8707090020179749, "num_tokens": 577347825.0, "step": 15137 }, { "epoch": 1.9257091973031422, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.140228271484375, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.869567334651947, "num_tokens": 577379981.0, "step": 15138 }, { "epoch": 1.9258364075817327, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.196649551391602, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8737752437591553, "num_tokens": 577414134.0, "step": 15139 }, { "epoch": 1.9259636178603232, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.155912399291992, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8679474592208862, "num_tokens": 577451170.0, "step": 15140 }, { "epoch": 1.9260908281389137, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.262121200561523, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8568979501724243, "num_tokens": 577487447.0, "step": 15141 }, { "epoch": 1.9262180384175043, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.16890811920166, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8741554617881775, "num_tokens": 577525732.0, "step": 15142 }, { "epoch": 1.9263452486960948, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.130083084106445, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8678728342056274, "num_tokens": 577563303.0, "step": 15143 }, { "epoch": 1.926472458974685, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.206514358520508, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8828675746917725, "num_tokens": 577607592.0, "step": 15144 }, { "epoch": 1.9265996692532756, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.173561096191406, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8615530729293823, "num_tokens": 577651646.0, "step": 15145 }, { "epoch": 1.9267268795318662, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.190159797668457, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8775092959403992, "num_tokens": 577689530.0, "step": 15146 }, { "epoch": 1.9268540898104567, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.155712127685547, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8650254607200623, "num_tokens": 577732524.0, "step": 15147 }, { "epoch": 1.9269813000890472, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.269875526428223, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8738157749176025, "num_tokens": 577767164.0, "step": 15148 }, { "epoch": 1.9271085103676378, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.15394401550293, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8763529658317566, "num_tokens": 577802347.0, "step": 15149 }, { "epoch": 1.927235720646228, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.284506797790527, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8614369630813599, "num_tokens": 577842899.0, "step": 15150 }, { "epoch": 1.9273629309248186, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.149059295654297, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8690940737724304, "num_tokens": 577885137.0, "step": 15151 }, { "epoch": 1.9274901412034091, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.222867012023926, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8829041123390198, "num_tokens": 577925727.0, "step": 15152 }, { "epoch": 1.9276173514819996, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.230316162109375, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8729496002197266, "num_tokens": 577965399.0, "step": 15153 }, { "epoch": 1.9277445617605902, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.192605972290039, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8697938919067383, "num_tokens": 578007608.0, "step": 15154 }, { "epoch": 1.9278717720391807, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.229684829711914, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8809508681297302, "num_tokens": 578043425.0, "step": 15155 }, { "epoch": 1.9279989823177712, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.320817947387695, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8670708537101746, "num_tokens": 578075838.0, "step": 15156 }, { "epoch": 1.9281261925963618, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.240995407104492, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8647208213806152, "num_tokens": 578110113.0, "step": 15157 }, { "epoch": 1.9282534028749523, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.272350311279297, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8699290752410889, "num_tokens": 578145743.0, "step": 15158 }, { "epoch": 1.9283806131535428, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.214762687683105, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8779269456863403, "num_tokens": 578181588.0, "step": 15159 }, { "epoch": 1.9285078234321333, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.313400268554688, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8674428462982178, "num_tokens": 578222048.0, "step": 15160 }, { "epoch": 1.9286350337107239, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.21367359161377, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8719042539596558, "num_tokens": 578262664.0, "step": 15161 }, { "epoch": 1.9287622439893144, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.237738609313965, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8584129810333252, "num_tokens": 578303431.0, "step": 15162 }, { "epoch": 1.928889454267905, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.291448593139648, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8598660230636597, "num_tokens": 578338978.0, "step": 15163 }, { "epoch": 1.9290166645464955, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.251191139221191, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8661153316497803, "num_tokens": 578376340.0, "step": 15164 }, { "epoch": 1.929143874825086, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.253993034362793, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8711785078048706, "num_tokens": 578414714.0, "step": 15165 }, { "epoch": 1.9292710851036765, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.17914867401123, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8565444946289062, "num_tokens": 578453668.0, "step": 15166 }, { "epoch": 1.929398295382267, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.242226600646973, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8623970150947571, "num_tokens": 578494148.0, "step": 15167 }, { "epoch": 1.9295255056608573, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.22280216217041, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8489397764205933, "num_tokens": 578530743.0, "step": 15168 }, { "epoch": 1.9296527159394479, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.266375541687012, "learning_rate": 1e-06, "loss": 0.3435, "mean_token_accuracy": 0.8917245864868164, "num_tokens": 578567613.0, "step": 15169 }, { "epoch": 1.9297799262180384, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.235466957092285, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8675712943077087, "num_tokens": 578602981.0, "step": 15170 }, { "epoch": 1.929907136496629, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.214622497558594, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8630821108818054, "num_tokens": 578641027.0, "step": 15171 }, { "epoch": 1.9300343467752195, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.204323768615723, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8825567960739136, "num_tokens": 578683006.0, "step": 15172 }, { "epoch": 1.93016155705381, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.24339485168457, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8737778663635254, "num_tokens": 578716722.0, "step": 15173 }, { "epoch": 1.9302887673324005, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.231492042541504, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8505207896232605, "num_tokens": 578750740.0, "step": 15174 }, { "epoch": 1.9304159776109908, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.188155174255371, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8723577260971069, "num_tokens": 578791358.0, "step": 15175 }, { "epoch": 1.9305431878895813, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.245901107788086, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8534693717956543, "num_tokens": 578834514.0, "step": 15176 }, { "epoch": 1.9306703981681719, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.218863487243652, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8612802624702454, "num_tokens": 578875869.0, "step": 15177 }, { "epoch": 1.9307976084467624, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.191910743713379, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8735095858573914, "num_tokens": 578921079.0, "step": 15178 }, { "epoch": 1.930924818725353, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.22556209564209, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.868586540222168, "num_tokens": 578959837.0, "step": 15179 }, { "epoch": 1.9310520290039435, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.186505317687988, "learning_rate": 1e-06, "loss": 0.5102, "mean_token_accuracy": 0.8413642644882202, "num_tokens": 579001629.0, "step": 15180 }, { "epoch": 1.931179239282534, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.235544204711914, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8920621871948242, "num_tokens": 579036762.0, "step": 15181 }, { "epoch": 1.9313064495611245, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.184104919433594, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8474367260932922, "num_tokens": 579078455.0, "step": 15182 }, { "epoch": 1.931433659839715, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.23177433013916, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8726562261581421, "num_tokens": 579116942.0, "step": 15183 }, { "epoch": 1.9315608701183056, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.240181922912598, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8816092014312744, "num_tokens": 579155180.0, "step": 15184 }, { "epoch": 1.931688080396896, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.285947799682617, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8620076775550842, "num_tokens": 579194336.0, "step": 15185 }, { "epoch": 1.9318152906754866, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.255306243896484, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8660241365432739, "num_tokens": 579226609.0, "step": 15186 }, { "epoch": 1.9319425009540772, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.274575233459473, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8725866675376892, "num_tokens": 579259288.0, "step": 15187 }, { "epoch": 1.9320697112326677, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.236392974853516, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8775538206100464, "num_tokens": 579294797.0, "step": 15188 }, { "epoch": 1.9321969215112582, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.21993637084961, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8802651762962341, "num_tokens": 579334352.0, "step": 15189 }, { "epoch": 1.9323241317898487, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.252306938171387, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8585993051528931, "num_tokens": 579379618.0, "step": 15190 }, { "epoch": 1.9324513420684393, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.24814224243164, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8687323331832886, "num_tokens": 579420230.0, "step": 15191 }, { "epoch": 1.9325785523470298, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.272157669067383, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8698585033416748, "num_tokens": 579458236.0, "step": 15192 }, { "epoch": 1.93270576262562, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.300440788269043, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8681714534759521, "num_tokens": 579495561.0, "step": 15193 }, { "epoch": 1.9328329729042106, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.244009017944336, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8680772185325623, "num_tokens": 579530225.0, "step": 15194 }, { "epoch": 1.9329601831828012, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.22135066986084, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8609136939048767, "num_tokens": 579570143.0, "step": 15195 }, { "epoch": 1.9330873934613917, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.254688262939453, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8815642595291138, "num_tokens": 579607105.0, "step": 15196 }, { "epoch": 1.9332146037399822, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.191617965698242, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8839367628097534, "num_tokens": 579640018.0, "step": 15197 }, { "epoch": 1.9333418140185727, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.190437316894531, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8642162084579468, "num_tokens": 579681250.0, "step": 15198 }, { "epoch": 1.933469024297163, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.258408546447754, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8582248687744141, "num_tokens": 579718535.0, "step": 15199 }, { "epoch": 1.9335962345757536, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.2281494140625, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8590033650398254, "num_tokens": 579756959.0, "step": 15200 }, { "epoch": 1.933723444854344, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.326488494873047, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8599706888198853, "num_tokens": 579798266.0, "step": 15201 }, { "epoch": 1.9338506551329346, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.227499008178711, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8884791135787964, "num_tokens": 579831869.0, "step": 15202 }, { "epoch": 1.9339778654115252, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.294648170471191, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.87090003490448, "num_tokens": 579874168.0, "step": 15203 }, { "epoch": 1.9341050756901157, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.223709106445312, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8756516575813293, "num_tokens": 579916313.0, "step": 15204 }, { "epoch": 1.9342322859687062, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.276137351989746, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8809762001037598, "num_tokens": 579959022.0, "step": 15205 }, { "epoch": 1.9343594962472968, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.256706237792969, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8552292585372925, "num_tokens": 580003675.0, "step": 15206 }, { "epoch": 1.9344867065258873, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.232016563415527, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8783528804779053, "num_tokens": 580044384.0, "step": 15207 }, { "epoch": 1.9346139168044778, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.341019630432129, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8589088916778564, "num_tokens": 580078955.0, "step": 15208 }, { "epoch": 1.9347411270830683, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.20995044708252, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.865776777267456, "num_tokens": 580116793.0, "step": 15209 }, { "epoch": 1.9348683373616589, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.212532997131348, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.868083655834198, "num_tokens": 580153622.0, "step": 15210 }, { "epoch": 1.9349955476402494, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.265568733215332, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8671002388000488, "num_tokens": 580195287.0, "step": 15211 }, { "epoch": 1.93512275791884, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.248924255371094, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8574718236923218, "num_tokens": 580229601.0, "step": 15212 }, { "epoch": 1.9352499681974304, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.24670696258545, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8714169263839722, "num_tokens": 580268079.0, "step": 15213 }, { "epoch": 1.935377178476021, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.21091365814209, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8512146472930908, "num_tokens": 580305379.0, "step": 15214 }, { "epoch": 1.9355043887546115, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.248268127441406, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8691864013671875, "num_tokens": 580338438.0, "step": 15215 }, { "epoch": 1.935631599033202, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.171695709228516, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8695926666259766, "num_tokens": 580370017.0, "step": 15216 }, { "epoch": 1.9357588093117923, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.202664375305176, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8793116807937622, "num_tokens": 580409572.0, "step": 15217 }, { "epoch": 1.9358860195903829, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.145647048950195, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8745419979095459, "num_tokens": 580448481.0, "step": 15218 }, { "epoch": 1.9360132298689734, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.219347953796387, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8812230825424194, "num_tokens": 580481919.0, "step": 15219 }, { "epoch": 1.936140440147564, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.201006889343262, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8643935918807983, "num_tokens": 580520758.0, "step": 15220 }, { "epoch": 1.9362676504261545, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.23299503326416, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8597106337547302, "num_tokens": 580559839.0, "step": 15221 }, { "epoch": 1.936394860704745, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.229148864746094, "learning_rate": 1e-06, "loss": 0.3355, "mean_token_accuracy": 0.8961095809936523, "num_tokens": 580591767.0, "step": 15222 }, { "epoch": 1.9365220709833355, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.166171073913574, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8774406313896179, "num_tokens": 580633180.0, "step": 15223 }, { "epoch": 1.9366492812619258, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.19416618347168, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8720467686653137, "num_tokens": 580677045.0, "step": 15224 }, { "epoch": 1.9367764915405163, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.228622436523438, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8826786875724792, "num_tokens": 580717423.0, "step": 15225 }, { "epoch": 1.9369037018191069, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.165057182312012, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8741669654846191, "num_tokens": 580756677.0, "step": 15226 }, { "epoch": 1.9370309120976974, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.202313423156738, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8746373057365417, "num_tokens": 580796993.0, "step": 15227 }, { "epoch": 1.937158122376288, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.208162307739258, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8730310201644897, "num_tokens": 580834616.0, "step": 15228 }, { "epoch": 1.9372853326548785, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.228616714477539, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8721070885658264, "num_tokens": 580873718.0, "step": 15229 }, { "epoch": 1.937412542933469, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.198813438415527, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8653663396835327, "num_tokens": 580911678.0, "step": 15230 }, { "epoch": 1.9375397532120595, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.227684020996094, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8744446039199829, "num_tokens": 580953282.0, "step": 15231 }, { "epoch": 1.93766696349065, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.11679744720459, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8753425478935242, "num_tokens": 580997954.0, "step": 15232 }, { "epoch": 1.9377941737692406, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.243073463439941, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.870743989944458, "num_tokens": 581035623.0, "step": 15233 }, { "epoch": 1.937921384047831, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.165306091308594, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8786036968231201, "num_tokens": 581068727.0, "step": 15234 }, { "epoch": 1.9380485943264216, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.217628479003906, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.867673933506012, "num_tokens": 581111180.0, "step": 15235 }, { "epoch": 1.9381758046050122, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.18667984008789, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8497233390808105, "num_tokens": 581152373.0, "step": 15236 }, { "epoch": 1.9383030148836027, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.193142890930176, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8680782318115234, "num_tokens": 581195854.0, "step": 15237 }, { "epoch": 1.9384302251621932, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.212946891784668, "learning_rate": 1e-06, "loss": 0.5007, "mean_token_accuracy": 0.8428869247436523, "num_tokens": 581236414.0, "step": 15238 }, { "epoch": 1.9385574354407837, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.170266151428223, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8532557487487793, "num_tokens": 581275349.0, "step": 15239 }, { "epoch": 1.9386846457193743, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.249785423278809, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8774619102478027, "num_tokens": 581313231.0, "step": 15240 }, { "epoch": 1.9388118559979648, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.170939445495605, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8710537552833557, "num_tokens": 581355295.0, "step": 15241 }, { "epoch": 1.938939066276555, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.181207656860352, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8730183839797974, "num_tokens": 581392633.0, "step": 15242 }, { "epoch": 1.9390662765551456, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.266252517700195, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8831727504730225, "num_tokens": 581425552.0, "step": 15243 }, { "epoch": 1.9391934868337362, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.243097305297852, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8589649200439453, "num_tokens": 581467707.0, "step": 15244 }, { "epoch": 1.9393206971123267, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.18342399597168, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.875450611114502, "num_tokens": 581504888.0, "step": 15245 }, { "epoch": 1.9394479073909172, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.206123352050781, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8699455261230469, "num_tokens": 581537800.0, "step": 15246 }, { "epoch": 1.9395751176695077, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.265182495117188, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8906472325325012, "num_tokens": 581570594.0, "step": 15247 }, { "epoch": 1.939702327948098, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.195026397705078, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8634870648384094, "num_tokens": 581610764.0, "step": 15248 }, { "epoch": 1.9398295382266886, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.186939239501953, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8698410987854004, "num_tokens": 581652341.0, "step": 15249 }, { "epoch": 1.939956748505279, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.250353813171387, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8612918853759766, "num_tokens": 581689978.0, "step": 15250 }, { "epoch": 1.9400839587838696, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.224695205688477, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8785440921783447, "num_tokens": 581728811.0, "step": 15251 }, { "epoch": 1.9402111690624602, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.203129768371582, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8775653839111328, "num_tokens": 581768016.0, "step": 15252 }, { "epoch": 1.9403383793410507, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.268686294555664, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8720205426216125, "num_tokens": 581805405.0, "step": 15253 }, { "epoch": 1.9404655896196412, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.301645278930664, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8514857292175293, "num_tokens": 581845302.0, "step": 15254 }, { "epoch": 1.9405927998982317, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.282387733459473, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8786134719848633, "num_tokens": 581875391.0, "step": 15255 }, { "epoch": 1.9407200101768223, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.290300369262695, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8687899708747864, "num_tokens": 581916580.0, "step": 15256 }, { "epoch": 1.9408472204554128, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.163328170776367, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8705971837043762, "num_tokens": 581958539.0, "step": 15257 }, { "epoch": 1.9409744307340033, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.294638633728027, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8690904974937439, "num_tokens": 581996981.0, "step": 15258 }, { "epoch": 1.9411016410125939, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.281635284423828, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8791455626487732, "num_tokens": 582036068.0, "step": 15259 }, { "epoch": 1.9412288512911844, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.249504089355469, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8723764419555664, "num_tokens": 582073322.0, "step": 15260 }, { "epoch": 1.941356061569775, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.199423789978027, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8724997639656067, "num_tokens": 582110736.0, "step": 15261 }, { "epoch": 1.9414832718483654, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.240434646606445, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8537759780883789, "num_tokens": 582153431.0, "step": 15262 }, { "epoch": 1.941610482126956, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.237751960754395, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8516746163368225, "num_tokens": 582191793.0, "step": 15263 }, { "epoch": 1.9417376924055465, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.18152141571045, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8604150414466858, "num_tokens": 582227923.0, "step": 15264 }, { "epoch": 1.941864902684137, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.22372817993164, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8556582927703857, "num_tokens": 582268029.0, "step": 15265 }, { "epoch": 1.9419921129627273, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.179085731506348, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8647361397743225, "num_tokens": 582302173.0, "step": 15266 }, { "epoch": 1.9421193232413179, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.190512657165527, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8842750787734985, "num_tokens": 582341642.0, "step": 15267 }, { "epoch": 1.9422465335199084, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.24937915802002, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8724145889282227, "num_tokens": 582378960.0, "step": 15268 }, { "epoch": 1.942373743798499, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.204914093017578, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8767037987709045, "num_tokens": 582422149.0, "step": 15269 }, { "epoch": 1.9425009540770894, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.245017051696777, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.853614091873169, "num_tokens": 582462657.0, "step": 15270 }, { "epoch": 1.94262816435568, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.2080078125, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8823598623275757, "num_tokens": 582499264.0, "step": 15271 }, { "epoch": 1.9427553746342705, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.258339881896973, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8730559349060059, "num_tokens": 582534744.0, "step": 15272 }, { "epoch": 1.9428825849128608, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.204721450805664, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8660717606544495, "num_tokens": 582577010.0, "step": 15273 }, { "epoch": 1.9430097951914513, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.316526412963867, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8653551340103149, "num_tokens": 582617103.0, "step": 15274 }, { "epoch": 1.9431370054700419, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.266341209411621, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8641079664230347, "num_tokens": 582661754.0, "step": 15275 }, { "epoch": 1.9432642157486324, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.231389045715332, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8715171813964844, "num_tokens": 582701701.0, "step": 15276 }, { "epoch": 1.943391426027223, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.243148803710938, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8606873750686646, "num_tokens": 582741977.0, "step": 15277 }, { "epoch": 1.9435186363058135, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.196234703063965, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8789637684822083, "num_tokens": 582783420.0, "step": 15278 }, { "epoch": 1.943645846584404, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.24863052368164, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.875028133392334, "num_tokens": 582819464.0, "step": 15279 }, { "epoch": 1.9437730568629945, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.25793170928955, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8754871487617493, "num_tokens": 582860757.0, "step": 15280 }, { "epoch": 1.943900267141585, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.237004280090332, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8650063276290894, "num_tokens": 582910300.0, "step": 15281 }, { "epoch": 1.9440274774201756, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.25848388671875, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8626088500022888, "num_tokens": 582956333.0, "step": 15282 }, { "epoch": 1.944154687698766, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.234844207763672, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8782820701599121, "num_tokens": 582994068.0, "step": 15283 }, { "epoch": 1.9442818979773566, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.251646995544434, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8677220344543457, "num_tokens": 583027829.0, "step": 15284 }, { "epoch": 1.9444091082559471, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.219686508178711, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8676128387451172, "num_tokens": 583068881.0, "step": 15285 }, { "epoch": 1.9445363185345377, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.264572143554688, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.873448371887207, "num_tokens": 583100097.0, "step": 15286 }, { "epoch": 1.9446635288131282, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.266022682189941, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8690529465675354, "num_tokens": 583135340.0, "step": 15287 }, { "epoch": 1.9447907390917187, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.18209171295166, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8690012693405151, "num_tokens": 583172032.0, "step": 15288 }, { "epoch": 1.9449179493703093, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.265880584716797, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8742161989212036, "num_tokens": 583205244.0, "step": 15289 }, { "epoch": 1.9450451596488998, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.249789237976074, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8660751581192017, "num_tokens": 583247537.0, "step": 15290 }, { "epoch": 1.94517236992749, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.288838386535645, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8631459474563599, "num_tokens": 583285349.0, "step": 15291 }, { "epoch": 1.9452995802060806, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.278255462646484, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8720625042915344, "num_tokens": 583323810.0, "step": 15292 }, { "epoch": 1.9454267904846712, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.310937881469727, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8824625015258789, "num_tokens": 583354244.0, "step": 15293 }, { "epoch": 1.9455540007632617, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.200778007507324, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8835017681121826, "num_tokens": 583391764.0, "step": 15294 }, { "epoch": 1.9456812110418522, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.266596794128418, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8566136956214905, "num_tokens": 583426305.0, "step": 15295 }, { "epoch": 1.9458084213204427, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.183475494384766, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8668074607849121, "num_tokens": 583465252.0, "step": 15296 }, { "epoch": 1.945935631599033, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.355706214904785, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8614790439605713, "num_tokens": 583510001.0, "step": 15297 }, { "epoch": 1.9460628418776236, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.160454750061035, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8715213537216187, "num_tokens": 583547380.0, "step": 15298 }, { "epoch": 1.946190052156214, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.305059432983398, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8704166412353516, "num_tokens": 583588852.0, "step": 15299 }, { "epoch": 1.9463172624348046, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.224908828735352, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8771868944168091, "num_tokens": 583634006.0, "step": 15300 }, { "epoch": 1.9464444727133952, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.269824981689453, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8561481833457947, "num_tokens": 583679430.0, "step": 15301 }, { "epoch": 1.9465716829919857, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.330586433410645, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8760614395141602, "num_tokens": 583708856.0, "step": 15302 }, { "epoch": 1.9466988932705762, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.28231143951416, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8681627511978149, "num_tokens": 583750737.0, "step": 15303 }, { "epoch": 1.9468261035491667, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.35861873626709, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8742567896842957, "num_tokens": 583786342.0, "step": 15304 }, { "epoch": 1.9469533138277573, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.247848510742188, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8757933378219604, "num_tokens": 583821681.0, "step": 15305 }, { "epoch": 1.9470805241063478, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.336175918579102, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8526602983474731, "num_tokens": 583855640.0, "step": 15306 }, { "epoch": 1.9472077343849383, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.30789852142334, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8797149658203125, "num_tokens": 583888047.0, "step": 15307 }, { "epoch": 1.9473349446635289, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.19843864440918, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8686270713806152, "num_tokens": 583923220.0, "step": 15308 }, { "epoch": 1.9474621549421194, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.279974937438965, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8648844957351685, "num_tokens": 583965718.0, "step": 15309 }, { "epoch": 1.94758936522071, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.262805938720703, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8750776052474976, "num_tokens": 584014027.0, "step": 15310 }, { "epoch": 1.9477165754993004, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.238813400268555, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8732851147651672, "num_tokens": 584045019.0, "step": 15311 }, { "epoch": 1.947843785777891, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.275157928466797, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8539009094238281, "num_tokens": 584086839.0, "step": 15312 }, { "epoch": 1.9479709960564815, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.258071899414062, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8840283751487732, "num_tokens": 584123785.0, "step": 15313 }, { "epoch": 1.948098206335072, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.286539077758789, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8614776730537415, "num_tokens": 584163768.0, "step": 15314 }, { "epoch": 1.9482254166136623, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.302253723144531, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8662496209144592, "num_tokens": 584201668.0, "step": 15315 }, { "epoch": 1.9483526268922529, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.225325584411621, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.878068208694458, "num_tokens": 584236626.0, "step": 15316 }, { "epoch": 1.9484798371708434, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.28560733795166, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8538216948509216, "num_tokens": 584279811.0, "step": 15317 }, { "epoch": 1.948607047449434, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.269951820373535, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8697165846824646, "num_tokens": 584317314.0, "step": 15318 }, { "epoch": 1.9487342577280244, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.272063255310059, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8754181861877441, "num_tokens": 584355768.0, "step": 15319 }, { "epoch": 1.948861468006615, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.213685989379883, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8670324087142944, "num_tokens": 584386205.0, "step": 15320 }, { "epoch": 1.9489886782852053, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.277255058288574, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8816169500350952, "num_tokens": 584419721.0, "step": 15321 }, { "epoch": 1.9491158885637958, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.269740104675293, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.867439866065979, "num_tokens": 584451029.0, "step": 15322 }, { "epoch": 1.9492430988423863, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.273832321166992, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8845319747924805, "num_tokens": 584483991.0, "step": 15323 }, { "epoch": 1.9493703091209769, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.260770797729492, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8683969974517822, "num_tokens": 584517652.0, "step": 15324 }, { "epoch": 1.9494975193995674, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.217743873596191, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8780725002288818, "num_tokens": 584561058.0, "step": 15325 }, { "epoch": 1.949624729678158, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.278934478759766, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8737035393714905, "num_tokens": 584595097.0, "step": 15326 }, { "epoch": 1.9497519399567484, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.192441940307617, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8772113919258118, "num_tokens": 584639073.0, "step": 15327 }, { "epoch": 1.949879150235339, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.278264045715332, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.86040198802948, "num_tokens": 584680517.0, "step": 15328 }, { "epoch": 1.9500063605139295, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.24563217163086, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8698187470436096, "num_tokens": 584717796.0, "step": 15329 }, { "epoch": 1.95013357079252, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.285558700561523, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8829594850540161, "num_tokens": 584755595.0, "step": 15330 }, { "epoch": 1.9502607810711106, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.273650169372559, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8596559762954712, "num_tokens": 584798681.0, "step": 15331 }, { "epoch": 1.950387991349701, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.282419204711914, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8731502890586853, "num_tokens": 584835424.0, "step": 15332 }, { "epoch": 1.9505152016282916, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.204764366149902, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8840388655662537, "num_tokens": 584872063.0, "step": 15333 }, { "epoch": 1.9506424119068821, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.291387557983398, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8853599429130554, "num_tokens": 584905850.0, "step": 15334 }, { "epoch": 1.9507696221854727, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.257439613342285, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8788006901741028, "num_tokens": 584946480.0, "step": 15335 }, { "epoch": 1.9508968324640632, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.25830364227295, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8765016198158264, "num_tokens": 584977382.0, "step": 15336 }, { "epoch": 1.9510240427426537, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.274115562438965, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8724423050880432, "num_tokens": 585010885.0, "step": 15337 }, { "epoch": 1.9511512530212443, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.348758697509766, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8686554431915283, "num_tokens": 585048208.0, "step": 15338 }, { "epoch": 1.9512784632998348, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.27229118347168, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8654924035072327, "num_tokens": 585087477.0, "step": 15339 }, { "epoch": 1.951405673578425, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.252120971679688, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8700353503227234, "num_tokens": 585120860.0, "step": 15340 }, { "epoch": 1.9515328838570156, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.250961303710938, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8713511228561401, "num_tokens": 585161887.0, "step": 15341 }, { "epoch": 1.9516600941356061, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.269404411315918, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8794413805007935, "num_tokens": 585200496.0, "step": 15342 }, { "epoch": 1.9517873044141967, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.362060546875, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8709452152252197, "num_tokens": 585243976.0, "step": 15343 }, { "epoch": 1.9519145146927872, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.217329978942871, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8652276992797852, "num_tokens": 585282366.0, "step": 15344 }, { "epoch": 1.9520417249713777, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.30656909942627, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8686047792434692, "num_tokens": 585325224.0, "step": 15345 }, { "epoch": 1.952168935249968, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.317262649536133, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8791525363922119, "num_tokens": 585361954.0, "step": 15346 }, { "epoch": 1.9522961455285586, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.279735565185547, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8674470782279968, "num_tokens": 585406033.0, "step": 15347 }, { "epoch": 1.952423355807149, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.278097152709961, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8688303232192993, "num_tokens": 585440457.0, "step": 15348 }, { "epoch": 1.9525505660857396, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.342517852783203, "learning_rate": 1e-06, "loss": 0.346, "mean_token_accuracy": 0.892715334892273, "num_tokens": 585473451.0, "step": 15349 }, { "epoch": 1.9526777763643302, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.255789756774902, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8739296197891235, "num_tokens": 585509652.0, "step": 15350 }, { "epoch": 1.9528049866429207, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.3085355758667, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8747360110282898, "num_tokens": 585545591.0, "step": 15351 }, { "epoch": 1.9529321969215112, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.249937057495117, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.859240710735321, "num_tokens": 585583757.0, "step": 15352 }, { "epoch": 1.9530594072001017, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.304734230041504, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8587642908096313, "num_tokens": 585623897.0, "step": 15353 }, { "epoch": 1.9531866174786923, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.255766868591309, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8744125366210938, "num_tokens": 585661043.0, "step": 15354 }, { "epoch": 1.9533138277572828, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.306918144226074, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8438286781311035, "num_tokens": 585699201.0, "step": 15355 }, { "epoch": 1.9534410380358733, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.276411056518555, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8631584644317627, "num_tokens": 585735325.0, "step": 15356 }, { "epoch": 1.9535682483144639, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.262779235839844, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8781634569168091, "num_tokens": 585769909.0, "step": 15357 }, { "epoch": 1.9536954585930544, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.330811500549316, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8581414222717285, "num_tokens": 585804590.0, "step": 15358 }, { "epoch": 1.953822668871645, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.272322654724121, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8657158613204956, "num_tokens": 585843349.0, "step": 15359 }, { "epoch": 1.9539498791502354, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.26001262664795, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.882570743560791, "num_tokens": 585881846.0, "step": 15360 }, { "epoch": 1.954077089428826, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.265698432922363, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8679362535476685, "num_tokens": 585917848.0, "step": 15361 }, { "epoch": 1.9542042997074165, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.232159614562988, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8759945631027222, "num_tokens": 585961249.0, "step": 15362 }, { "epoch": 1.954331509986007, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.334419250488281, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8782976865768433, "num_tokens": 586005795.0, "step": 15363 }, { "epoch": 1.9544587202645973, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.214289665222168, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8794603943824768, "num_tokens": 586041523.0, "step": 15364 }, { "epoch": 1.9545859305431879, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.303446769714355, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8616563081741333, "num_tokens": 586077800.0, "step": 15365 }, { "epoch": 1.9547131408217784, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.347208023071289, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8657655715942383, "num_tokens": 586116010.0, "step": 15366 }, { "epoch": 1.954840351100369, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.233146667480469, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8820359110832214, "num_tokens": 586155359.0, "step": 15367 }, { "epoch": 1.9549675613789594, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.292043685913086, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8757476806640625, "num_tokens": 586188190.0, "step": 15368 }, { "epoch": 1.95509477165755, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.27562427520752, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8696222305297852, "num_tokens": 586225971.0, "step": 15369 }, { "epoch": 1.9552219819361403, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.2928466796875, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8662073016166687, "num_tokens": 586263295.0, "step": 15370 }, { "epoch": 1.9553491922147308, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.211408615112305, "learning_rate": 1e-06, "loss": 0.3384, "mean_token_accuracy": 0.8915215730667114, "num_tokens": 586293547.0, "step": 15371 }, { "epoch": 1.9554764024933213, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.31494426727295, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8662488460540771, "num_tokens": 586333180.0, "step": 15372 }, { "epoch": 1.9556036127719119, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.285303115844727, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8716268539428711, "num_tokens": 586368780.0, "step": 15373 }, { "epoch": 1.9557308230505024, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.337066650390625, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8551634550094604, "num_tokens": 586404177.0, "step": 15374 }, { "epoch": 1.955858033329093, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.289204597473145, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8552600145339966, "num_tokens": 586446917.0, "step": 15375 }, { "epoch": 1.9559852436076834, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.35807991027832, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8711949586868286, "num_tokens": 586485832.0, "step": 15376 }, { "epoch": 1.956112453886274, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.271716117858887, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8805931806564331, "num_tokens": 586524089.0, "step": 15377 }, { "epoch": 1.9562396641648645, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.281994819641113, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8735525608062744, "num_tokens": 586564097.0, "step": 15378 }, { "epoch": 1.956366874443455, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.280532836914062, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8837562799453735, "num_tokens": 586600653.0, "step": 15379 }, { "epoch": 1.9564940847220456, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.321884155273438, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8764303922653198, "num_tokens": 586639527.0, "step": 15380 }, { "epoch": 1.956621295000636, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.305570602416992, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8757394552230835, "num_tokens": 586670885.0, "step": 15381 }, { "epoch": 1.9567485052792266, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.31551456451416, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8831475973129272, "num_tokens": 586709875.0, "step": 15382 }, { "epoch": 1.9568757155578171, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.266359329223633, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.864491879940033, "num_tokens": 586753164.0, "step": 15383 }, { "epoch": 1.9570029258364077, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.292332649230957, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8772104978561401, "num_tokens": 586793737.0, "step": 15384 }, { "epoch": 1.9571301361149982, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.23096752166748, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8670461773872375, "num_tokens": 586833428.0, "step": 15385 }, { "epoch": 1.9572573463935887, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.2680025100708, "learning_rate": 1e-06, "loss": 0.4862, "mean_token_accuracy": 0.8482970595359802, "num_tokens": 586871202.0, "step": 15386 }, { "epoch": 1.9573845566721793, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.242729187011719, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.881562352180481, "num_tokens": 586905537.0, "step": 15387 }, { "epoch": 1.9575117669507698, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.309216499328613, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8611454963684082, "num_tokens": 586940378.0, "step": 15388 }, { "epoch": 1.95763897722936, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.285945892333984, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8811691999435425, "num_tokens": 586978806.0, "step": 15389 }, { "epoch": 1.9577661875079506, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.301187515258789, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8649594783782959, "num_tokens": 587007367.0, "step": 15390 }, { "epoch": 1.9578933977865411, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.354843139648438, "learning_rate": 1e-06, "loss": 0.342, "mean_token_accuracy": 0.8921273350715637, "num_tokens": 587040422.0, "step": 15391 }, { "epoch": 1.9580206080651317, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.282442092895508, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8877363204956055, "num_tokens": 587070238.0, "step": 15392 }, { "epoch": 1.9581478183437222, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.29885482788086, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8771845102310181, "num_tokens": 587105973.0, "step": 15393 }, { "epoch": 1.9582750286223127, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.29628849029541, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8837888836860657, "num_tokens": 587137975.0, "step": 15394 }, { "epoch": 1.958402238900903, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.285689353942871, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.867352306842804, "num_tokens": 587169176.0, "step": 15395 }, { "epoch": 1.9585294491794936, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.274879455566406, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8682237863540649, "num_tokens": 587205570.0, "step": 15396 }, { "epoch": 1.958656659458084, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.25786304473877, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.862002968788147, "num_tokens": 587246729.0, "step": 15397 }, { "epoch": 1.9587838697366746, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.32817554473877, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8755423426628113, "num_tokens": 587284537.0, "step": 15398 }, { "epoch": 1.9589110800152651, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.281363487243652, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8763620853424072, "num_tokens": 587324394.0, "step": 15399 }, { "epoch": 1.9590382902938557, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.348698616027832, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8782522082328796, "num_tokens": 587359935.0, "step": 15400 }, { "epoch": 1.9591655005724462, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.224369049072266, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8826203346252441, "num_tokens": 587402081.0, "step": 15401 }, { "epoch": 1.9592927108510367, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.249812126159668, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8678991794586182, "num_tokens": 587442939.0, "step": 15402 }, { "epoch": 1.9594199211296273, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.296314239501953, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.86269211769104, "num_tokens": 587479155.0, "step": 15403 }, { "epoch": 1.9595471314082178, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.338705062866211, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8600547313690186, "num_tokens": 587519285.0, "step": 15404 }, { "epoch": 1.9596743416868083, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.229666709899902, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8534026741981506, "num_tokens": 587563978.0, "step": 15405 }, { "epoch": 1.9598015519653988, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.316601753234863, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8650882840156555, "num_tokens": 587599052.0, "step": 15406 }, { "epoch": 1.9599287622439894, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.28909683227539, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8740701079368591, "num_tokens": 587639980.0, "step": 15407 }, { "epoch": 1.96005597252258, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.256369590759277, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8670939207077026, "num_tokens": 587673836.0, "step": 15408 }, { "epoch": 1.9601831828011704, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.28842830657959, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8714056015014648, "num_tokens": 587712896.0, "step": 15409 }, { "epoch": 1.960310393079761, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.28128433227539, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8647835850715637, "num_tokens": 587747352.0, "step": 15410 }, { "epoch": 1.9604376033583515, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.20199966430664, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8698224425315857, "num_tokens": 587790234.0, "step": 15411 }, { "epoch": 1.960564813636942, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.324012756347656, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8844078779220581, "num_tokens": 587830185.0, "step": 15412 }, { "epoch": 1.9606920239155323, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.251530647277832, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8723562359809875, "num_tokens": 587866265.0, "step": 15413 }, { "epoch": 1.9608192341941229, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.250725746154785, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8830620050430298, "num_tokens": 587904956.0, "step": 15414 }, { "epoch": 1.9609464444727134, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.302008628845215, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8780174851417542, "num_tokens": 587942369.0, "step": 15415 }, { "epoch": 1.961073654751304, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.265238761901855, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8757599592208862, "num_tokens": 587979340.0, "step": 15416 }, { "epoch": 1.9612008650298944, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.30579948425293, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8666700124740601, "num_tokens": 588021812.0, "step": 15417 }, { "epoch": 1.961328075308485, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.267081260681152, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8674838542938232, "num_tokens": 588064978.0, "step": 15418 }, { "epoch": 1.9614552855870753, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.285024642944336, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8663259744644165, "num_tokens": 588098142.0, "step": 15419 }, { "epoch": 1.9615824958656658, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.258426666259766, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8781176805496216, "num_tokens": 588141808.0, "step": 15420 }, { "epoch": 1.9617097061442563, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.27151107788086, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8782480955123901, "num_tokens": 588178137.0, "step": 15421 }, { "epoch": 1.9618369164228469, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.329615592956543, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8651599884033203, "num_tokens": 588215646.0, "step": 15422 }, { "epoch": 1.9619641267014374, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.203536033630371, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8618549704551697, "num_tokens": 588251345.0, "step": 15423 }, { "epoch": 1.962091336980028, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.255728721618652, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8771897554397583, "num_tokens": 588291222.0, "step": 15424 }, { "epoch": 1.9622185472586184, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.266240119934082, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8630353212356567, "num_tokens": 588329480.0, "step": 15425 }, { "epoch": 1.962345757537209, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.204399108886719, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8619884848594666, "num_tokens": 588367143.0, "step": 15426 }, { "epoch": 1.9624729678157995, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.257733345031738, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8790664076805115, "num_tokens": 588408363.0, "step": 15427 }, { "epoch": 1.96260017809439, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.154084205627441, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8774999380111694, "num_tokens": 588445397.0, "step": 15428 }, { "epoch": 1.9627273883729806, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.263772964477539, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8765048980712891, "num_tokens": 588480436.0, "step": 15429 }, { "epoch": 1.962854598651571, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.265578269958496, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8695026636123657, "num_tokens": 588526918.0, "step": 15430 }, { "epoch": 1.9629818089301616, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.225849151611328, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8715975880622864, "num_tokens": 588561300.0, "step": 15431 }, { "epoch": 1.9631090192087521, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.301396369934082, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8614028096199036, "num_tokens": 588603393.0, "step": 15432 }, { "epoch": 1.9632362294873427, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.25093936920166, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8786007165908813, "num_tokens": 588643806.0, "step": 15433 }, { "epoch": 1.9633634397659332, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.192956924438477, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8699764013290405, "num_tokens": 588685090.0, "step": 15434 }, { "epoch": 1.9634906500445237, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.308365821838379, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8597460985183716, "num_tokens": 588720983.0, "step": 15435 }, { "epoch": 1.9636178603231143, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.237207412719727, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8643779754638672, "num_tokens": 588760143.0, "step": 15436 }, { "epoch": 1.9637450706017048, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.278115272521973, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8734797239303589, "num_tokens": 588796636.0, "step": 15437 }, { "epoch": 1.963872280880295, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.29234790802002, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8680189251899719, "num_tokens": 588831681.0, "step": 15438 }, { "epoch": 1.9639994911588856, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.2429838180542, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8812963962554932, "num_tokens": 588875408.0, "step": 15439 }, { "epoch": 1.9641267014374761, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.242112159729004, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8751459121704102, "num_tokens": 588913583.0, "step": 15440 }, { "epoch": 1.9642539117160667, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.269776344299316, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8795501589775085, "num_tokens": 588949833.0, "step": 15441 }, { "epoch": 1.9643811219946572, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.320536613464355, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.876142680644989, "num_tokens": 588985678.0, "step": 15442 }, { "epoch": 1.9645083322732477, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.270764350891113, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8696267604827881, "num_tokens": 589018276.0, "step": 15443 }, { "epoch": 1.964635542551838, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.267773628234863, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8693997263908386, "num_tokens": 589061900.0, "step": 15444 }, { "epoch": 1.9647627528304286, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.27215576171875, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.873742938041687, "num_tokens": 589102740.0, "step": 15445 }, { "epoch": 1.964889963109019, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.305436134338379, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8800031542778015, "num_tokens": 589135276.0, "step": 15446 }, { "epoch": 1.9650171733876096, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.286792755126953, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.87163245677948, "num_tokens": 589173390.0, "step": 15447 }, { "epoch": 1.9651443836662001, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.322608947753906, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8643008470535278, "num_tokens": 589214592.0, "step": 15448 }, { "epoch": 1.9652715939447907, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.295184135437012, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8720047473907471, "num_tokens": 589250682.0, "step": 15449 }, { "epoch": 1.9653988042233812, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.319653511047363, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8652321100234985, "num_tokens": 589288326.0, "step": 15450 }, { "epoch": 1.9655260145019717, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.334066390991211, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8677359819412231, "num_tokens": 589321632.0, "step": 15451 }, { "epoch": 1.9656532247805623, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.263627052307129, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8621256351470947, "num_tokens": 589361134.0, "step": 15452 }, { "epoch": 1.9657804350591528, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.260560035705566, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8732213973999023, "num_tokens": 589402338.0, "step": 15453 }, { "epoch": 1.9659076453377433, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.274629592895508, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8650627136230469, "num_tokens": 589441080.0, "step": 15454 }, { "epoch": 1.9660348556163338, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.260960578918457, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8682472705841064, "num_tokens": 589484019.0, "step": 15455 }, { "epoch": 1.9661620658949244, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.23947525024414, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.865983247756958, "num_tokens": 589525890.0, "step": 15456 }, { "epoch": 1.966289276173515, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.346297264099121, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8761855959892273, "num_tokens": 589560644.0, "step": 15457 }, { "epoch": 1.9664164864521054, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.265052795410156, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8907395601272583, "num_tokens": 589600863.0, "step": 15458 }, { "epoch": 1.966543696730696, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.30128002166748, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8706991672515869, "num_tokens": 589637948.0, "step": 15459 }, { "epoch": 1.9666709070092865, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.370336532592773, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.883755624294281, "num_tokens": 589678061.0, "step": 15460 }, { "epoch": 1.966798117287877, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.31475830078125, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8598883152008057, "num_tokens": 589718050.0, "step": 15461 }, { "epoch": 1.9669253275664673, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.313187599182129, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8737956881523132, "num_tokens": 589759206.0, "step": 15462 }, { "epoch": 1.9670525378450578, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.304941177368164, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8589849472045898, "num_tokens": 589800218.0, "step": 15463 }, { "epoch": 1.9671797481236484, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.330109596252441, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8720920085906982, "num_tokens": 589843105.0, "step": 15464 }, { "epoch": 1.967306958402239, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.291189193725586, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8625858426094055, "num_tokens": 589881560.0, "step": 15465 }, { "epoch": 1.9674341686808294, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.280830383300781, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8590167760848999, "num_tokens": 589922755.0, "step": 15466 }, { "epoch": 1.96756137895942, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.43700122833252, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8521929979324341, "num_tokens": 589964982.0, "step": 15467 }, { "epoch": 1.9676885892380103, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.295190811157227, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8574330806732178, "num_tokens": 590003371.0, "step": 15468 }, { "epoch": 1.9678157995166008, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.307249069213867, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8684072494506836, "num_tokens": 590048361.0, "step": 15469 }, { "epoch": 1.9679430097951913, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.373754501342773, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8685557842254639, "num_tokens": 590083141.0, "step": 15470 }, { "epoch": 1.9680702200737819, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.216740608215332, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.864761471748352, "num_tokens": 590128161.0, "step": 15471 }, { "epoch": 1.9681974303523724, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.299357414245605, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.885496199131012, "num_tokens": 590166101.0, "step": 15472 }, { "epoch": 1.968324640630963, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.227155685424805, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.869675874710083, "num_tokens": 590207643.0, "step": 15473 }, { "epoch": 1.9684518509095534, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.315741539001465, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8630988001823425, "num_tokens": 590247567.0, "step": 15474 }, { "epoch": 1.968579061188144, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.247737884521484, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8660769462585449, "num_tokens": 590287940.0, "step": 15475 }, { "epoch": 1.9687062714667345, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.292011260986328, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8645000457763672, "num_tokens": 590322558.0, "step": 15476 }, { "epoch": 1.968833481745325, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.277132987976074, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.869505763053894, "num_tokens": 590362913.0, "step": 15477 }, { "epoch": 1.9689606920239155, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.243560791015625, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8619204759597778, "num_tokens": 590405877.0, "step": 15478 }, { "epoch": 1.969087902302506, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.337546348571777, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8612015247344971, "num_tokens": 590445373.0, "step": 15479 }, { "epoch": 1.9692151125810966, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.231966972351074, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8597428798675537, "num_tokens": 590483319.0, "step": 15480 }, { "epoch": 1.9693423228596871, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.248980522155762, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8568910956382751, "num_tokens": 590522433.0, "step": 15481 }, { "epoch": 1.9694695331382777, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.266983032226562, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8742760419845581, "num_tokens": 590564342.0, "step": 15482 }, { "epoch": 1.9695967434168682, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.351194381713867, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8733416795730591, "num_tokens": 590594527.0, "step": 15483 }, { "epoch": 1.9697239536954587, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.31418514251709, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8815518617630005, "num_tokens": 590630573.0, "step": 15484 }, { "epoch": 1.9698511639740492, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.37235164642334, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8711514472961426, "num_tokens": 590669333.0, "step": 15485 }, { "epoch": 1.9699783742526398, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.249322891235352, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8865176439285278, "num_tokens": 590709884.0, "step": 15486 }, { "epoch": 1.97010558453123, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.296441078186035, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8761717081069946, "num_tokens": 590746109.0, "step": 15487 }, { "epoch": 1.9702327948098206, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.341055870056152, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8550388813018799, "num_tokens": 590787881.0, "step": 15488 }, { "epoch": 1.9703600050884111, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.278897285461426, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8563448786735535, "num_tokens": 590826861.0, "step": 15489 }, { "epoch": 1.9704872153670017, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.317666053771973, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.878441572189331, "num_tokens": 590863104.0, "step": 15490 }, { "epoch": 1.9706144256455922, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.246414184570312, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8771281242370605, "num_tokens": 590900582.0, "step": 15491 }, { "epoch": 1.9707416359241827, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.370838165283203, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8772636651992798, "num_tokens": 590937735.0, "step": 15492 }, { "epoch": 1.970868846202773, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.269545555114746, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8850763440132141, "num_tokens": 590973089.0, "step": 15493 }, { "epoch": 1.9709960564813636, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.265718460083008, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8568280935287476, "num_tokens": 591012117.0, "step": 15494 }, { "epoch": 1.971123266759954, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.30188274383545, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8514782190322876, "num_tokens": 591044978.0, "step": 15495 }, { "epoch": 1.9712504770385446, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.218677520751953, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8674643039703369, "num_tokens": 591085562.0, "step": 15496 }, { "epoch": 1.9713776873171351, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.269688606262207, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8610069155693054, "num_tokens": 591120140.0, "step": 15497 }, { "epoch": 1.9715048975957257, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.334641456604004, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8610941767692566, "num_tokens": 591153952.0, "step": 15498 }, { "epoch": 1.9716321078743162, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.237161636352539, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.886329710483551, "num_tokens": 591192983.0, "step": 15499 }, { "epoch": 1.9717593181529067, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.2930269241333, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8712859153747559, "num_tokens": 591226260.0, "step": 15500 }, { "epoch": 1.9718865284314973, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.312007904052734, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8569738864898682, "num_tokens": 591259361.0, "step": 15501 }, { "epoch": 1.9720137387100878, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.24172306060791, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8711796402931213, "num_tokens": 591299753.0, "step": 15502 }, { "epoch": 1.9721409489886783, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.297059059143066, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8726949691772461, "num_tokens": 591335008.0, "step": 15503 }, { "epoch": 1.9722681592672688, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.338495254516602, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8760694861412048, "num_tokens": 591373001.0, "step": 15504 }, { "epoch": 1.9723953695458594, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.20930004119873, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8726168870925903, "num_tokens": 591408637.0, "step": 15505 }, { "epoch": 1.97252257982445, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.313021659851074, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8672405481338501, "num_tokens": 591444908.0, "step": 15506 }, { "epoch": 1.9726497901030404, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.278382301330566, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8732826709747314, "num_tokens": 591484641.0, "step": 15507 }, { "epoch": 1.972777000381631, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.256847381591797, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8673738837242126, "num_tokens": 591521859.0, "step": 15508 }, { "epoch": 1.9729042106602215, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.32380199432373, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.869512677192688, "num_tokens": 591559124.0, "step": 15509 }, { "epoch": 1.973031420938812, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.250639915466309, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8682954907417297, "num_tokens": 591597170.0, "step": 15510 }, { "epoch": 1.9731586312174023, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.339641571044922, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.875648021697998, "num_tokens": 591634784.0, "step": 15511 }, { "epoch": 1.9732858414959928, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.224994659423828, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8745203018188477, "num_tokens": 591671904.0, "step": 15512 }, { "epoch": 1.9734130517745834, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.314208984375, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8715211153030396, "num_tokens": 591712990.0, "step": 15513 }, { "epoch": 1.973540262053174, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.279755592346191, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8673762083053589, "num_tokens": 591748579.0, "step": 15514 }, { "epoch": 1.9736674723317644, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.222970008850098, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8638741970062256, "num_tokens": 591785865.0, "step": 15515 }, { "epoch": 1.973794682610355, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.337739944458008, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8794263601303101, "num_tokens": 591820087.0, "step": 15516 }, { "epoch": 1.9739218928889453, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.26527214050293, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8694199323654175, "num_tokens": 591859713.0, "step": 15517 }, { "epoch": 1.9740491031675358, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.327997207641602, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8778244256973267, "num_tokens": 591893142.0, "step": 15518 }, { "epoch": 1.9741763134461263, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.347235679626465, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8657158613204956, "num_tokens": 591931529.0, "step": 15519 }, { "epoch": 1.9743035237247168, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.312272071838379, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8753312826156616, "num_tokens": 591976248.0, "step": 15520 }, { "epoch": 1.9744307340033074, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.32837200164795, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8679962158203125, "num_tokens": 592018008.0, "step": 15521 }, { "epoch": 1.974557944281898, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.323528289794922, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8550714254379272, "num_tokens": 592062934.0, "step": 15522 }, { "epoch": 1.9746851545604884, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.29892349243164, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8688631057739258, "num_tokens": 592103129.0, "step": 15523 }, { "epoch": 1.974812364839079, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.319082260131836, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8651756048202515, "num_tokens": 592142116.0, "step": 15524 }, { "epoch": 1.9749395751176695, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.340974807739258, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8705222606658936, "num_tokens": 592182507.0, "step": 15525 }, { "epoch": 1.97506678539626, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.293034553527832, "learning_rate": 1e-06, "loss": 0.4724, "mean_token_accuracy": 0.8489861488342285, "num_tokens": 592225593.0, "step": 15526 }, { "epoch": 1.9751939956748505, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.310992240905762, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.894068717956543, "num_tokens": 592270676.0, "step": 15527 }, { "epoch": 1.975321205953441, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.295696258544922, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.888508141040802, "num_tokens": 592305537.0, "step": 15528 }, { "epoch": 1.9754484162320316, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.263545989990234, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8773118853569031, "num_tokens": 592338289.0, "step": 15529 }, { "epoch": 1.9755756265106221, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.273839950561523, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.873298704624176, "num_tokens": 592379189.0, "step": 15530 }, { "epoch": 1.9757028367892127, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.234031677246094, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.884352445602417, "num_tokens": 592421875.0, "step": 15531 }, { "epoch": 1.9758300470678032, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.288764953613281, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.851777195930481, "num_tokens": 592457438.0, "step": 15532 }, { "epoch": 1.9759572573463937, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.278111457824707, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8812752962112427, "num_tokens": 592493230.0, "step": 15533 }, { "epoch": 1.9760844676249842, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.198081016540527, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8545050024986267, "num_tokens": 592533656.0, "step": 15534 }, { "epoch": 1.9762116779035748, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.288692474365234, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.883264422416687, "num_tokens": 592571622.0, "step": 15535 }, { "epoch": 1.976338888182165, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.247639656066895, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8651878237724304, "num_tokens": 592613260.0, "step": 15536 }, { "epoch": 1.9764660984607556, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.337642669677734, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8674370050430298, "num_tokens": 592648872.0, "step": 15537 }, { "epoch": 1.9765933087393461, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.321634292602539, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8714087009429932, "num_tokens": 592682336.0, "step": 15538 }, { "epoch": 1.9767205190179367, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.310591697692871, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8842118382453918, "num_tokens": 592716412.0, "step": 15539 }, { "epoch": 1.9768477292965272, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.329042434692383, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8734800815582275, "num_tokens": 592753936.0, "step": 15540 }, { "epoch": 1.9769749395751177, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.225090980529785, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8704575300216675, "num_tokens": 592792731.0, "step": 15541 }, { "epoch": 1.977102149853708, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.267606735229492, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8737826347351074, "num_tokens": 592832035.0, "step": 15542 }, { "epoch": 1.9772293601322986, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.36660385131836, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8881579637527466, "num_tokens": 592869693.0, "step": 15543 }, { "epoch": 1.977356570410889, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.238457679748535, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8737409710884094, "num_tokens": 592909648.0, "step": 15544 }, { "epoch": 1.9774837806894796, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.333293914794922, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.874001145362854, "num_tokens": 592942539.0, "step": 15545 }, { "epoch": 1.9776109909680701, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.275818824768066, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8696157932281494, "num_tokens": 592983680.0, "step": 15546 }, { "epoch": 1.9777382012466607, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.31040096282959, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8600407242774963, "num_tokens": 593023161.0, "step": 15547 }, { "epoch": 1.9778654115252512, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.2880277633667, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8523290753364563, "num_tokens": 593066612.0, "step": 15548 }, { "epoch": 1.9779926218038417, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.265783309936523, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8587031960487366, "num_tokens": 593105224.0, "step": 15549 }, { "epoch": 1.9781198320824323, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.337665557861328, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.870914876461029, "num_tokens": 593143683.0, "step": 15550 }, { "epoch": 1.9782470423610228, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.325273513793945, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8562591075897217, "num_tokens": 593179827.0, "step": 15551 }, { "epoch": 1.9783742526396133, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.310030937194824, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8710941076278687, "num_tokens": 593220936.0, "step": 15552 }, { "epoch": 1.9785014629182038, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.284351348876953, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8659661412239075, "num_tokens": 593258503.0, "step": 15553 }, { "epoch": 1.9786286731967944, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.341438293457031, "learning_rate": 1e-06, "loss": 0.4927, "mean_token_accuracy": 0.8512809872627258, "num_tokens": 593293738.0, "step": 15554 }, { "epoch": 1.978755883475385, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.342012405395508, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8611681461334229, "num_tokens": 593338037.0, "step": 15555 }, { "epoch": 1.9788830937539754, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.342615127563477, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8544738292694092, "num_tokens": 593383509.0, "step": 15556 }, { "epoch": 1.979010304032566, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.282508850097656, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8740842342376709, "num_tokens": 593417734.0, "step": 15557 }, { "epoch": 1.9791375143111565, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.272201538085938, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8829174637794495, "num_tokens": 593454502.0, "step": 15558 }, { "epoch": 1.979264724589747, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.230905532836914, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8666737675666809, "num_tokens": 593498864.0, "step": 15559 }, { "epoch": 1.9793919348683373, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.312403678894043, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8665416836738586, "num_tokens": 593540125.0, "step": 15560 }, { "epoch": 1.9795191451469278, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.340985298156738, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8803306818008423, "num_tokens": 593566937.0, "step": 15561 }, { "epoch": 1.9796463554255184, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.249035835266113, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.866153359413147, "num_tokens": 593600109.0, "step": 15562 }, { "epoch": 1.979773565704109, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.19172477722168, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8688085079193115, "num_tokens": 593638982.0, "step": 15563 }, { "epoch": 1.9799007759826994, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.313631057739258, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8605935573577881, "num_tokens": 593677270.0, "step": 15564 }, { "epoch": 1.98002798626129, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.255921363830566, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8775869607925415, "num_tokens": 593714056.0, "step": 15565 }, { "epoch": 1.9801551965398803, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.2938814163208, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8602148294448853, "num_tokens": 593753516.0, "step": 15566 }, { "epoch": 1.9802824068184708, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.255106925964355, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.872148871421814, "num_tokens": 593789276.0, "step": 15567 }, { "epoch": 1.9804096170970613, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.246452331542969, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8746554851531982, "num_tokens": 593828112.0, "step": 15568 }, { "epoch": 1.9805368273756518, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.284871101379395, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8586207628250122, "num_tokens": 593866370.0, "step": 15569 }, { "epoch": 1.9806640376542424, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.245908737182617, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8746557235717773, "num_tokens": 593907863.0, "step": 15570 }, { "epoch": 1.980791247932833, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.2806396484375, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8811575770378113, "num_tokens": 593945254.0, "step": 15571 }, { "epoch": 1.9809184582114234, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.274580001831055, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8733503222465515, "num_tokens": 593987105.0, "step": 15572 }, { "epoch": 1.981045668490014, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.354443550109863, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8814481496810913, "num_tokens": 594024091.0, "step": 15573 }, { "epoch": 1.9811728787686045, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.285345077514648, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8805664777755737, "num_tokens": 594056018.0, "step": 15574 }, { "epoch": 1.981300089047195, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.26858901977539, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8725959658622742, "num_tokens": 594102297.0, "step": 15575 }, { "epoch": 1.9814272993257855, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.228428840637207, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8705629110336304, "num_tokens": 594140278.0, "step": 15576 }, { "epoch": 1.981554509604376, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.36187744140625, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.874417781829834, "num_tokens": 594175025.0, "step": 15577 }, { "epoch": 1.9816817198829666, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.22348403930664, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8690803050994873, "num_tokens": 594217906.0, "step": 15578 }, { "epoch": 1.9818089301615571, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.320154190063477, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.852825403213501, "num_tokens": 594264331.0, "step": 15579 }, { "epoch": 1.9819361404401477, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.27418041229248, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8856430053710938, "num_tokens": 594300220.0, "step": 15580 }, { "epoch": 1.9820633507187382, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.337244033813477, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.877094030380249, "num_tokens": 594337849.0, "step": 15581 }, { "epoch": 1.9821905609973287, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.224625587463379, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8645853996276855, "num_tokens": 594379295.0, "step": 15582 }, { "epoch": 1.9823177712759192, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.322454452514648, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8669113516807556, "num_tokens": 594408350.0, "step": 15583 }, { "epoch": 1.9824449815545098, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.262795448303223, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8593897819519043, "num_tokens": 594455110.0, "step": 15584 }, { "epoch": 1.9825721918331, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.23030948638916, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8749094009399414, "num_tokens": 594499231.0, "step": 15585 }, { "epoch": 1.9826994021116906, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.3098783493042, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8514590263366699, "num_tokens": 594541478.0, "step": 15586 }, { "epoch": 1.9828266123902811, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.32813549041748, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8730183839797974, "num_tokens": 594579075.0, "step": 15587 }, { "epoch": 1.9829538226688717, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.277010917663574, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8868299722671509, "num_tokens": 594616189.0, "step": 15588 }, { "epoch": 1.9830810329474622, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.357834815979004, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8774620294570923, "num_tokens": 594656633.0, "step": 15589 }, { "epoch": 1.9832082432260527, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.41209602355957, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8518601059913635, "num_tokens": 594696833.0, "step": 15590 }, { "epoch": 1.983335453504643, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.281950950622559, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8584630489349365, "num_tokens": 594735138.0, "step": 15591 }, { "epoch": 1.9834626637832335, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.288369178771973, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8642295002937317, "num_tokens": 594769587.0, "step": 15592 }, { "epoch": 1.983589874061824, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.401254653930664, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.876210629940033, "num_tokens": 594803635.0, "step": 15593 }, { "epoch": 1.9837170843404146, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.254843711853027, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8863472938537598, "num_tokens": 594839872.0, "step": 15594 }, { "epoch": 1.9838442946190051, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.301207542419434, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.858360767364502, "num_tokens": 594879208.0, "step": 15595 }, { "epoch": 1.9839715048975957, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.269003868103027, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8768817186355591, "num_tokens": 594910377.0, "step": 15596 }, { "epoch": 1.9840987151761862, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.315913200378418, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8728072643280029, "num_tokens": 594956574.0, "step": 15597 }, { "epoch": 1.9842259254547767, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.250523567199707, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8549156188964844, "num_tokens": 594998036.0, "step": 15598 }, { "epoch": 1.9843531357333672, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.345388412475586, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8651319146156311, "num_tokens": 595031404.0, "step": 15599 }, { "epoch": 1.9844803460119578, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.317280769348145, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8787131309509277, "num_tokens": 595069800.0, "step": 15600 }, { "epoch": 1.9846075562905483, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.293086051940918, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.875446081161499, "num_tokens": 595110987.0, "step": 15601 }, { "epoch": 1.9847347665691388, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.310728073120117, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8734560012817383, "num_tokens": 595149192.0, "step": 15602 }, { "epoch": 1.9848619768477294, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.309765815734863, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8694140911102295, "num_tokens": 595182993.0, "step": 15603 }, { "epoch": 1.9849891871263199, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.278142929077148, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8676128387451172, "num_tokens": 595216177.0, "step": 15604 }, { "epoch": 1.9851163974049104, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.318304061889648, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8717041015625, "num_tokens": 595253379.0, "step": 15605 }, { "epoch": 1.985243607683501, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.218841552734375, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8751450777053833, "num_tokens": 595292878.0, "step": 15606 }, { "epoch": 1.9853708179620915, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.343605041503906, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8713722229003906, "num_tokens": 595329082.0, "step": 15607 }, { "epoch": 1.985498028240682, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.337606430053711, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8634066581726074, "num_tokens": 595367246.0, "step": 15608 }, { "epoch": 1.9856252385192723, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.2620267868042, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8771238923072815, "num_tokens": 595404450.0, "step": 15609 }, { "epoch": 1.9857524487978628, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.378369331359863, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8829222917556763, "num_tokens": 595439718.0, "step": 15610 }, { "epoch": 1.9858796590764534, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.322944641113281, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8727861046791077, "num_tokens": 595473617.0, "step": 15611 }, { "epoch": 1.986006869355044, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.288336753845215, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8682375550270081, "num_tokens": 595516432.0, "step": 15612 }, { "epoch": 1.9861340796336344, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.238312721252441, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.871623158454895, "num_tokens": 595556278.0, "step": 15613 }, { "epoch": 1.986261289912225, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.338778495788574, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8760988712310791, "num_tokens": 595598713.0, "step": 15614 }, { "epoch": 1.9863885001908153, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.35677719116211, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8735255599021912, "num_tokens": 595630597.0, "step": 15615 }, { "epoch": 1.9865157104694058, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.296188354492188, "learning_rate": 1e-06, "loss": 0.3486, "mean_token_accuracy": 0.8884613513946533, "num_tokens": 595664163.0, "step": 15616 }, { "epoch": 1.9866429207479963, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.236347198486328, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8795329928398132, "num_tokens": 595698809.0, "step": 15617 }, { "epoch": 1.9867701310265868, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.310813903808594, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8655831217765808, "num_tokens": 595741166.0, "step": 15618 }, { "epoch": 1.9868973413051774, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.250502586364746, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8663740754127502, "num_tokens": 595787079.0, "step": 15619 }, { "epoch": 1.987024551583768, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.327167510986328, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8467584848403931, "num_tokens": 595829085.0, "step": 15620 }, { "epoch": 1.9871517618623584, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.256025314331055, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8686671853065491, "num_tokens": 595870646.0, "step": 15621 }, { "epoch": 1.987278972140949, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.261152267456055, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8602529764175415, "num_tokens": 595918473.0, "step": 15622 }, { "epoch": 1.9874061824195395, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.343822479248047, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8604649305343628, "num_tokens": 595954097.0, "step": 15623 }, { "epoch": 1.98753339269813, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.286908149719238, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8527772426605225, "num_tokens": 595990191.0, "step": 15624 }, { "epoch": 1.9876606029767205, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.270249366760254, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8637058734893799, "num_tokens": 596027109.0, "step": 15625 }, { "epoch": 1.987787813255311, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.259027481079102, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8788413405418396, "num_tokens": 596067930.0, "step": 15626 }, { "epoch": 1.9879150235339016, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.269161224365234, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8810498714447021, "num_tokens": 596105809.0, "step": 15627 }, { "epoch": 1.9880422338124921, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.3278169631958, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8740075826644897, "num_tokens": 596144220.0, "step": 15628 }, { "epoch": 1.9881694440910826, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.316883087158203, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8485154509544373, "num_tokens": 596183825.0, "step": 15629 }, { "epoch": 1.9882966543696732, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.241218566894531, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8788127899169922, "num_tokens": 596232333.0, "step": 15630 }, { "epoch": 1.9884238646482637, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.285093307495117, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8785660862922668, "num_tokens": 596272138.0, "step": 15631 }, { "epoch": 1.9885510749268542, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.259414672851562, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8800565004348755, "num_tokens": 596310254.0, "step": 15632 }, { "epoch": 1.9886782852054448, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.280084609985352, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.867495059967041, "num_tokens": 596350396.0, "step": 15633 }, { "epoch": 1.988805495484035, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.33729362487793, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8739259839057922, "num_tokens": 596388809.0, "step": 15634 }, { "epoch": 1.9889327057626256, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.260730743408203, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8782578706741333, "num_tokens": 596427366.0, "step": 15635 }, { "epoch": 1.9890599160412161, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.382278442382812, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8618919253349304, "num_tokens": 596469999.0, "step": 15636 }, { "epoch": 1.9891871263198067, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.346199989318848, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8560470342636108, "num_tokens": 596506773.0, "step": 15637 }, { "epoch": 1.9893143365983972, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.214632034301758, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8722086548805237, "num_tokens": 596547440.0, "step": 15638 }, { "epoch": 1.9894415468769877, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.409905433654785, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8772904276847839, "num_tokens": 596580212.0, "step": 15639 }, { "epoch": 1.989568757155578, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.243577003479004, "learning_rate": 1e-06, "loss": 0.345, "mean_token_accuracy": 0.8930955529212952, "num_tokens": 596611790.0, "step": 15640 }, { "epoch": 1.9896959674341685, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.312614440917969, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.8858605623245239, "num_tokens": 596646984.0, "step": 15641 }, { "epoch": 1.989823177712759, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.282755851745605, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8663183450698853, "num_tokens": 596688362.0, "step": 15642 }, { "epoch": 1.9899503879913496, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.354644775390625, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.873075008392334, "num_tokens": 596725298.0, "step": 15643 }, { "epoch": 1.9900775982699401, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.370616912841797, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8772070407867432, "num_tokens": 596757400.0, "step": 15644 }, { "epoch": 1.9902048085485307, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.2256498336792, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8827639818191528, "num_tokens": 596790533.0, "step": 15645 }, { "epoch": 1.9903320188271212, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.308005332946777, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8649183511734009, "num_tokens": 596834467.0, "step": 15646 }, { "epoch": 1.9904592291057117, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.295720100402832, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8586716651916504, "num_tokens": 596873262.0, "step": 15647 }, { "epoch": 1.9905864393843022, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.266563415527344, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8710103034973145, "num_tokens": 596912833.0, "step": 15648 }, { "epoch": 1.9907136496628928, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.269769668579102, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.878848135471344, "num_tokens": 596953554.0, "step": 15649 }, { "epoch": 1.9908408599414833, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.33241081237793, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8663219213485718, "num_tokens": 596991206.0, "step": 15650 }, { "epoch": 1.9909680702200738, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.264432907104492, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.872173011302948, "num_tokens": 597027222.0, "step": 15651 }, { "epoch": 1.9910952804986644, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.306192398071289, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8799887299537659, "num_tokens": 597069040.0, "step": 15652 }, { "epoch": 1.9912224907772549, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.28512954711914, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8628044128417969, "num_tokens": 597112529.0, "step": 15653 }, { "epoch": 1.9913497010558454, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.353647232055664, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8726987242698669, "num_tokens": 597148304.0, "step": 15654 }, { "epoch": 1.991476911334436, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.287643432617188, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8795846700668335, "num_tokens": 597184244.0, "step": 15655 }, { "epoch": 1.9916041216130265, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.32669448852539, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8638957738876343, "num_tokens": 597218178.0, "step": 15656 }, { "epoch": 1.991731331891617, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.295326232910156, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8686542510986328, "num_tokens": 597259256.0, "step": 15657 }, { "epoch": 1.9918585421702073, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.323366165161133, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8814342021942139, "num_tokens": 597296584.0, "step": 15658 }, { "epoch": 1.9919857524487978, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.361928939819336, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8570661544799805, "num_tokens": 597341116.0, "step": 15659 }, { "epoch": 1.9921129627273884, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.33449935913086, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8691070079803467, "num_tokens": 597375595.0, "step": 15660 }, { "epoch": 1.9922401730059789, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.345613479614258, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8573493361473083, "num_tokens": 597419226.0, "step": 15661 }, { "epoch": 1.9923673832845694, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.403242111206055, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8509333729743958, "num_tokens": 597461242.0, "step": 15662 }, { "epoch": 1.99249459356316, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.301002502441406, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8703820705413818, "num_tokens": 597500030.0, "step": 15663 }, { "epoch": 1.9926218038417502, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.357436180114746, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8559678196907043, "num_tokens": 597540428.0, "step": 15664 }, { "epoch": 1.9927490141203408, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.31671142578125, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8808401823043823, "num_tokens": 597575070.0, "step": 15665 }, { "epoch": 1.9928762243989313, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.27248764038086, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8687736392021179, "num_tokens": 597617425.0, "step": 15666 }, { "epoch": 1.9930034346775218, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.365225791931152, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8775131702423096, "num_tokens": 597660489.0, "step": 15667 }, { "epoch": 1.9931306449561124, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.299015998840332, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8763253688812256, "num_tokens": 597703626.0, "step": 15668 }, { "epoch": 1.993257855234703, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.285368919372559, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.866813600063324, "num_tokens": 597741176.0, "step": 15669 }, { "epoch": 1.9933850655132934, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.3933744430542, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8584252595901489, "num_tokens": 597779763.0, "step": 15670 }, { "epoch": 1.993512275791884, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.339555740356445, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8731663227081299, "num_tokens": 597813300.0, "step": 15671 }, { "epoch": 1.9936394860704745, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.382725715637207, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8602396845817566, "num_tokens": 597849062.0, "step": 15672 }, { "epoch": 1.993766696349065, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.351469039916992, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.870569109916687, "num_tokens": 597886565.0, "step": 15673 }, { "epoch": 1.9938939066276555, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.339296340942383, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.873413622379303, "num_tokens": 597925097.0, "step": 15674 }, { "epoch": 1.994021116906246, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.355695724487305, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8835117816925049, "num_tokens": 597960786.0, "step": 15675 }, { "epoch": 1.9941483271848366, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.30793285369873, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8519693613052368, "num_tokens": 598000442.0, "step": 15676 }, { "epoch": 1.9942755374634271, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.39604663848877, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8682608604431152, "num_tokens": 598042946.0, "step": 15677 }, { "epoch": 1.9944027477420176, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.314852714538574, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8503572940826416, "num_tokens": 598081659.0, "step": 15678 }, { "epoch": 1.9945299580206082, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.327123641967773, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8650531768798828, "num_tokens": 598120203.0, "step": 15679 }, { "epoch": 1.9946571682991987, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.374576568603516, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8580637574195862, "num_tokens": 598154779.0, "step": 15680 }, { "epoch": 1.9947843785777892, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.312114715576172, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8705232739448547, "num_tokens": 598197075.0, "step": 15681 }, { "epoch": 1.9949115888563798, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.267760276794434, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8625997304916382, "num_tokens": 598240539.0, "step": 15682 }, { "epoch": 1.99503879913497, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.357481956481934, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8722596168518066, "num_tokens": 598281776.0, "step": 15683 }, { "epoch": 1.9951660094135606, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.324616432189941, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8771259188652039, "num_tokens": 598322715.0, "step": 15684 }, { "epoch": 1.9952932196921511, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.332404136657715, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8649184703826904, "num_tokens": 598364495.0, "step": 15685 }, { "epoch": 1.9954204299707416, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.332682609558105, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8673661947250366, "num_tokens": 598407102.0, "step": 15686 }, { "epoch": 1.9955476402493322, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.313106536865234, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8572279810905457, "num_tokens": 598448600.0, "step": 15687 }, { "epoch": 1.9956748505279227, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.376953125, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8599340319633484, "num_tokens": 598491140.0, "step": 15688 }, { "epoch": 1.995802060806513, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.328070640563965, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8668656349182129, "num_tokens": 598527831.0, "step": 15689 }, { "epoch": 1.9959292710851035, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.341483116149902, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8568546772003174, "num_tokens": 598568680.0, "step": 15690 }, { "epoch": 1.996056481363694, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.30895709991455, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8594486713409424, "num_tokens": 598606254.0, "step": 15691 }, { "epoch": 1.9961836916422846, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.37982177734375, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8716399669647217, "num_tokens": 598642400.0, "step": 15692 }, { "epoch": 1.9963109019208751, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.283677101135254, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8484615683555603, "num_tokens": 598678379.0, "step": 15693 }, { "epoch": 1.9964381121994657, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.332747459411621, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8678802251815796, "num_tokens": 598718234.0, "step": 15694 }, { "epoch": 1.9965653224780562, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.317240715026855, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8726398348808289, "num_tokens": 598754897.0, "step": 15695 }, { "epoch": 1.9966925327566467, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.330500602722168, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8646144866943359, "num_tokens": 598785268.0, "step": 15696 }, { "epoch": 1.9968197430352372, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.278665542602539, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8758252859115601, "num_tokens": 598827131.0, "step": 15697 }, { "epoch": 1.9969469533138278, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.348104476928711, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8704212307929993, "num_tokens": 598859507.0, "step": 15698 }, { "epoch": 1.9970741635924183, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.340612411499023, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8600013256072998, "num_tokens": 598893280.0, "step": 15699 }, { "epoch": 1.9972013738710088, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.300559997558594, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8832802772521973, "num_tokens": 598932759.0, "step": 15700 }, { "epoch": 1.9973285841495994, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.384499549865723, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8797067403793335, "num_tokens": 598972899.0, "step": 15701 }, { "epoch": 1.9974557944281899, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.343029975891113, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8719224333763123, "num_tokens": 599012811.0, "step": 15702 }, { "epoch": 1.9975830047067804, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.263233184814453, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8605022430419922, "num_tokens": 599055632.0, "step": 15703 }, { "epoch": 1.997710214985371, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.357885360717773, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8517565727233887, "num_tokens": 599090924.0, "step": 15704 }, { "epoch": 1.9978374252639615, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.273420333862305, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8834729194641113, "num_tokens": 599130434.0, "step": 15705 }, { "epoch": 1.997964635542552, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.34032154083252, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.868598997592926, "num_tokens": 599171018.0, "step": 15706 }, { "epoch": 1.9980918458211423, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.302958488464355, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8611990213394165, "num_tokens": 599212445.0, "step": 15707 }, { "epoch": 1.9982190560997328, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.279953002929688, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.874126672744751, "num_tokens": 599250591.0, "step": 15708 }, { "epoch": 1.9983462663783234, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.333809852600098, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.865903377532959, "num_tokens": 599284987.0, "step": 15709 }, { "epoch": 1.9984734766569139, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.390710830688477, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8655486702919006, "num_tokens": 599324656.0, "step": 15710 }, { "epoch": 1.9986006869355044, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.319009780883789, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.879652738571167, "num_tokens": 599364014.0, "step": 15711 }, { "epoch": 1.998727897214095, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.43407917022705, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8614089488983154, "num_tokens": 599396781.0, "step": 15712 }, { "epoch": 1.9988551074926852, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.300514221191406, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8792768120765686, "num_tokens": 599433240.0, "step": 15713 }, { "epoch": 1.9989823177712758, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.374484062194824, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.879413366317749, "num_tokens": 599474883.0, "step": 15714 }, { "epoch": 1.9991095280498663, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.34216594696045, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8702448606491089, "num_tokens": 599508492.0, "step": 15715 }, { "epoch": 1.9992367383284568, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.398307800292969, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8722481727600098, "num_tokens": 599545946.0, "step": 15716 }, { "epoch": 1.9993639486070474, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.314255714416504, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8829584717750549, "num_tokens": 599586217.0, "step": 15717 }, { "epoch": 1.9994911588856379, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.362621307373047, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8767937421798706, "num_tokens": 599620431.0, "step": 15718 }, { "epoch": 1.9996183691642284, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.337799072265625, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8557461500167847, "num_tokens": 599659959.0, "step": 15719 }, { "epoch": 1.999745579442819, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.38017463684082, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8685367703437805, "num_tokens": 599696558.0, "step": 15720 }, { "epoch": 1.9998727897214095, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.373373031616211, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.859082043170929, "num_tokens": 599734925.0, "step": 15721 }, { "epoch": 2.0, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.358341217041016, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8738991022109985, "num_tokens": 599772613.0, "step": 15722 }, { "epoch": 2.0001272102785905, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.403255462646484, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8757453560829163, "num_tokens": 599813892.0, "step": 15723 }, { "epoch": 2.000254420557181, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.347783088684082, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8721495866775513, "num_tokens": 599852807.0, "step": 15724 }, { "epoch": 2.0003816308357716, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.416220664978027, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8788466453552246, "num_tokens": 599893712.0, "step": 15725 }, { "epoch": 2.000508841114362, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.334463119506836, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8736118674278259, "num_tokens": 599929954.0, "step": 15726 }, { "epoch": 2.0006360513929526, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.32969856262207, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8701328635215759, "num_tokens": 599967763.0, "step": 15727 }, { "epoch": 2.000763261671543, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.432555198669434, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.880537211894989, "num_tokens": 600008938.0, "step": 15728 }, { "epoch": 2.0008904719501337, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.341684341430664, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8683136701583862, "num_tokens": 600051849.0, "step": 15729 }, { "epoch": 2.0010176822287242, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.387335777282715, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8781413435935974, "num_tokens": 600088568.0, "step": 15730 }, { "epoch": 2.0011448925073148, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.3723726272583, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8765671849250793, "num_tokens": 600131553.0, "step": 15731 }, { "epoch": 2.0012721027859053, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.333093643188477, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8725854158401489, "num_tokens": 600173198.0, "step": 15732 }, { "epoch": 2.001399313064496, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.332569122314453, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8768471479415894, "num_tokens": 600211347.0, "step": 15733 }, { "epoch": 2.0015265233430863, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.256431579589844, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8667137622833252, "num_tokens": 600249953.0, "step": 15734 }, { "epoch": 2.0016537336216764, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.4058256149292, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8654038906097412, "num_tokens": 600291211.0, "step": 15735 }, { "epoch": 2.001780943900267, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.32041072845459, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8782336711883545, "num_tokens": 600332906.0, "step": 15736 }, { "epoch": 2.0019081541788575, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.331510543823242, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8734712600708008, "num_tokens": 600370570.0, "step": 15737 }, { "epoch": 2.002035364457448, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.409262657165527, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8635980486869812, "num_tokens": 600409368.0, "step": 15738 }, { "epoch": 2.0021625747360385, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.296777725219727, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8760499954223633, "num_tokens": 600442019.0, "step": 15739 }, { "epoch": 2.002289785014629, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.395608901977539, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8809404969215393, "num_tokens": 600471665.0, "step": 15740 }, { "epoch": 2.0024169952932196, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.303276062011719, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8713172674179077, "num_tokens": 600508804.0, "step": 15741 }, { "epoch": 2.00254420557181, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.34541130065918, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8615050315856934, "num_tokens": 600543895.0, "step": 15742 }, { "epoch": 2.0026714158504006, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.323442459106445, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8742220401763916, "num_tokens": 600585818.0, "step": 15743 }, { "epoch": 2.002798626128991, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.331888198852539, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8772450685501099, "num_tokens": 600623106.0, "step": 15744 }, { "epoch": 2.0029258364075817, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.336063385009766, "learning_rate": 1e-06, "loss": 0.3295, "mean_token_accuracy": 0.8933811187744141, "num_tokens": 600662662.0, "step": 15745 }, { "epoch": 2.0030530466861722, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.34913444519043, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8598077893257141, "num_tokens": 600699416.0, "step": 15746 }, { "epoch": 2.0031802569647628, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.373296737670898, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8662770986557007, "num_tokens": 600739351.0, "step": 15747 }, { "epoch": 2.0033074672433533, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.346992492675781, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8758293986320496, "num_tokens": 600781198.0, "step": 15748 }, { "epoch": 2.003434677521944, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.337163925170898, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8722866177558899, "num_tokens": 600821743.0, "step": 15749 }, { "epoch": 2.0035618878005343, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.351346015930176, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8879120945930481, "num_tokens": 600862463.0, "step": 15750 }, { "epoch": 2.003689098079125, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.436173439025879, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8630457520484924, "num_tokens": 600902629.0, "step": 15751 }, { "epoch": 2.0038163083577154, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.326638221740723, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8707705736160278, "num_tokens": 600941778.0, "step": 15752 }, { "epoch": 2.003943518636306, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.385587692260742, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.881074845790863, "num_tokens": 600979425.0, "step": 15753 }, { "epoch": 2.0040707289148965, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.383437156677246, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8688647150993347, "num_tokens": 601017882.0, "step": 15754 }, { "epoch": 2.004197939193487, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.300161361694336, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8677753210067749, "num_tokens": 601053394.0, "step": 15755 }, { "epoch": 2.0043251494720775, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.398245811462402, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8763883113861084, "num_tokens": 601094940.0, "step": 15756 }, { "epoch": 2.004452359750668, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.250163078308105, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8744833469390869, "num_tokens": 601132963.0, "step": 15757 }, { "epoch": 2.0045795700292586, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.367148399353027, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8718807697296143, "num_tokens": 601168511.0, "step": 15758 }, { "epoch": 2.0047067803078487, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.366389274597168, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.873847246170044, "num_tokens": 601208922.0, "step": 15759 }, { "epoch": 2.004833990586439, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.351485252380371, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.867387056350708, "num_tokens": 601253814.0, "step": 15760 }, { "epoch": 2.0049612008650297, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.31263256072998, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8740084767341614, "num_tokens": 601295584.0, "step": 15761 }, { "epoch": 2.0050884111436202, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.344478607177734, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8654043674468994, "num_tokens": 601329134.0, "step": 15762 }, { "epoch": 2.0052156214222108, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.373246192932129, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8812175393104553, "num_tokens": 601369100.0, "step": 15763 }, { "epoch": 2.0053428317008013, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.357893943786621, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.887548565864563, "num_tokens": 601407876.0, "step": 15764 }, { "epoch": 2.005470041979392, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.383889198303223, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8858797550201416, "num_tokens": 601446996.0, "step": 15765 }, { "epoch": 2.0055972522579824, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.353318214416504, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8649060726165771, "num_tokens": 601482158.0, "step": 15766 }, { "epoch": 2.005724462536573, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.402983665466309, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8783479928970337, "num_tokens": 601517940.0, "step": 15767 }, { "epoch": 2.0058516728151634, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.348155975341797, "learning_rate": 1e-06, "loss": 0.3291, "mean_token_accuracy": 0.8933393955230713, "num_tokens": 601553823.0, "step": 15768 }, { "epoch": 2.005978883093754, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.378527641296387, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8623273968696594, "num_tokens": 601595482.0, "step": 15769 }, { "epoch": 2.0061060933723445, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.352107048034668, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8819870352745056, "num_tokens": 601632451.0, "step": 15770 }, { "epoch": 2.006233303650935, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.36318302154541, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8713167309761047, "num_tokens": 601667573.0, "step": 15771 }, { "epoch": 2.0063605139295255, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.395285606384277, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8793481588363647, "num_tokens": 601704494.0, "step": 15772 }, { "epoch": 2.006487724208116, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.328579902648926, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8641392588615417, "num_tokens": 601745307.0, "step": 15773 }, { "epoch": 2.0066149344867066, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.379201889038086, "learning_rate": 1e-06, "loss": 0.35, "mean_token_accuracy": 0.8892456889152527, "num_tokens": 601782551.0, "step": 15774 }, { "epoch": 2.006742144765297, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.370514869689941, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8752971291542053, "num_tokens": 601823378.0, "step": 15775 }, { "epoch": 2.0068693550438876, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.348380088806152, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8744211792945862, "num_tokens": 601860881.0, "step": 15776 }, { "epoch": 2.006996565322478, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.324036598205566, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8718384504318237, "num_tokens": 601903502.0, "step": 15777 }, { "epoch": 2.0071237756010687, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.386125564575195, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8910598754882812, "num_tokens": 601937097.0, "step": 15778 }, { "epoch": 2.007250985879659, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.329826354980469, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8770760893821716, "num_tokens": 601978050.0, "step": 15779 }, { "epoch": 2.0073781961582498, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.368782997131348, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.871774435043335, "num_tokens": 602018167.0, "step": 15780 }, { "epoch": 2.0075054064368403, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.368152618408203, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8594163656234741, "num_tokens": 602054485.0, "step": 15781 }, { "epoch": 2.007632616715431, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.418630599975586, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8781931400299072, "num_tokens": 602092881.0, "step": 15782 }, { "epoch": 2.0077598269940213, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.362953186035156, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8749940395355225, "num_tokens": 602127424.0, "step": 15783 }, { "epoch": 2.0078870372726114, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.41226577758789, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.880616307258606, "num_tokens": 602168340.0, "step": 15784 }, { "epoch": 2.008014247551202, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.391797065734863, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8643958568572998, "num_tokens": 602207894.0, "step": 15785 }, { "epoch": 2.0081414578297925, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.376643180847168, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8761401176452637, "num_tokens": 602245924.0, "step": 15786 }, { "epoch": 2.008268668108383, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.419697761535645, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8718866109848022, "num_tokens": 602278417.0, "step": 15787 }, { "epoch": 2.0083958783869735, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.339608192443848, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8772258758544922, "num_tokens": 602318431.0, "step": 15788 }, { "epoch": 2.008523088665564, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.398813247680664, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8644762635231018, "num_tokens": 602356315.0, "step": 15789 }, { "epoch": 2.0086502989441546, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.328166007995605, "learning_rate": 1e-06, "loss": 0.3444, "mean_token_accuracy": 0.8885383605957031, "num_tokens": 602398865.0, "step": 15790 }, { "epoch": 2.008777509222745, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.377694129943848, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8846203684806824, "num_tokens": 602435835.0, "step": 15791 }, { "epoch": 2.0089047195013356, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.381494522094727, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8815937638282776, "num_tokens": 602470359.0, "step": 15792 }, { "epoch": 2.009031929779926, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.375292778015137, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8775573372840881, "num_tokens": 602512132.0, "step": 15793 }, { "epoch": 2.0091591400585167, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.338050842285156, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8711979389190674, "num_tokens": 602552700.0, "step": 15794 }, { "epoch": 2.0092863503371072, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.39047908782959, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8735111355781555, "num_tokens": 602587779.0, "step": 15795 }, { "epoch": 2.0094135606156978, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.357660293579102, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8641363382339478, "num_tokens": 602623834.0, "step": 15796 }, { "epoch": 2.0095407708942883, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.340503692626953, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8851256966590881, "num_tokens": 602660141.0, "step": 15797 }, { "epoch": 2.009667981172879, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.373627662658691, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8781377077102661, "num_tokens": 602698990.0, "step": 15798 }, { "epoch": 2.0097951914514693, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.455068588256836, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.855912983417511, "num_tokens": 602732053.0, "step": 15799 }, { "epoch": 2.00992240173006, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.379670143127441, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8751319050788879, "num_tokens": 602770975.0, "step": 15800 }, { "epoch": 2.0100496120086504, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.375718116760254, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8651515245437622, "num_tokens": 602813207.0, "step": 15801 }, { "epoch": 2.010176822287241, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.346278190612793, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8793729543685913, "num_tokens": 602849951.0, "step": 15802 }, { "epoch": 2.0103040325658315, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.410614013671875, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8652489185333252, "num_tokens": 602891418.0, "step": 15803 }, { "epoch": 2.010431242844422, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.479844093322754, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8798884749412537, "num_tokens": 602933039.0, "step": 15804 }, { "epoch": 2.0105584531230125, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.369033813476562, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.888647198677063, "num_tokens": 602965546.0, "step": 15805 }, { "epoch": 2.010685663401603, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.409845352172852, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8739216327667236, "num_tokens": 603004473.0, "step": 15806 }, { "epoch": 2.0108128736801936, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.363785743713379, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8669446110725403, "num_tokens": 603041644.0, "step": 15807 }, { "epoch": 2.0109400839587837, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.417184829711914, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8625244498252869, "num_tokens": 603084698.0, "step": 15808 }, { "epoch": 2.011067294237374, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.361762046813965, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8587838411331177, "num_tokens": 603121059.0, "step": 15809 }, { "epoch": 2.0111945045159647, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.388509750366211, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8714722394943237, "num_tokens": 603156604.0, "step": 15810 }, { "epoch": 2.0113217147945552, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.31485366821289, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8738410472869873, "num_tokens": 603183743.0, "step": 15811 }, { "epoch": 2.0114489250731458, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.389989852905273, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8510505557060242, "num_tokens": 603222099.0, "step": 15812 }, { "epoch": 2.0115761353517363, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.39593505859375, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8547500371932983, "num_tokens": 603263467.0, "step": 15813 }, { "epoch": 2.011703345630327, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.33584213256836, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8549290895462036, "num_tokens": 603296726.0, "step": 15814 }, { "epoch": 2.0118305559089174, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.36096477508545, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8791577219963074, "num_tokens": 603329691.0, "step": 15815 }, { "epoch": 2.011957766187508, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.344104766845703, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8813385963439941, "num_tokens": 603371619.0, "step": 15816 }, { "epoch": 2.0120849764660984, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.340921401977539, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.875351071357727, "num_tokens": 603400018.0, "step": 15817 }, { "epoch": 2.012212186744689, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.418597221374512, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8765379190444946, "num_tokens": 603436971.0, "step": 15818 }, { "epoch": 2.0123393970232795, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.34241771697998, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8744229078292847, "num_tokens": 603472802.0, "step": 15819 }, { "epoch": 2.01246660730187, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.439810752868652, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8795092105865479, "num_tokens": 603506547.0, "step": 15820 }, { "epoch": 2.0125938175804605, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.332598686218262, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8571134805679321, "num_tokens": 603544967.0, "step": 15821 }, { "epoch": 2.012721027859051, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.328411102294922, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8682411909103394, "num_tokens": 603582522.0, "step": 15822 }, { "epoch": 2.0128482381376416, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.385331153869629, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8825998902320862, "num_tokens": 603619584.0, "step": 15823 }, { "epoch": 2.012975448416232, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.328143119812012, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8615407943725586, "num_tokens": 603660542.0, "step": 15824 }, { "epoch": 2.0131026586948226, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.430736541748047, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.874848484992981, "num_tokens": 603698900.0, "step": 15825 }, { "epoch": 2.013229868973413, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.350926399230957, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8561556935310364, "num_tokens": 603733019.0, "step": 15826 }, { "epoch": 2.0133570792520037, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.31137752532959, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8826087117195129, "num_tokens": 603772071.0, "step": 15827 }, { "epoch": 2.013484289530594, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.406522750854492, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8659765720367432, "num_tokens": 603811008.0, "step": 15828 }, { "epoch": 2.0136114998091847, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.401708602905273, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8640849590301514, "num_tokens": 603852647.0, "step": 15829 }, { "epoch": 2.0137387100877753, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.33553695678711, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8724079132080078, "num_tokens": 603894198.0, "step": 15830 }, { "epoch": 2.013865920366366, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.39642333984375, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8700543642044067, "num_tokens": 603930168.0, "step": 15831 }, { "epoch": 2.0139931306449563, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.37852954864502, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8827621936798096, "num_tokens": 603966189.0, "step": 15832 }, { "epoch": 2.0141203409235464, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.346955299377441, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8693536520004272, "num_tokens": 604011110.0, "step": 15833 }, { "epoch": 2.014247551202137, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.366479873657227, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8560009002685547, "num_tokens": 604044256.0, "step": 15834 }, { "epoch": 2.0143747614807275, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.351541519165039, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8810322284698486, "num_tokens": 604078652.0, "step": 15835 }, { "epoch": 2.014501971759318, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.3551607131958, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8817601799964905, "num_tokens": 604112302.0, "step": 15836 }, { "epoch": 2.0146291820379085, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.34757137298584, "learning_rate": 1e-06, "loss": 0.3455, "mean_token_accuracy": 0.8892844915390015, "num_tokens": 604150011.0, "step": 15837 }, { "epoch": 2.014756392316499, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.452730178833008, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8733309507369995, "num_tokens": 604188859.0, "step": 15838 }, { "epoch": 2.0148836025950896, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.365909576416016, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8684402704238892, "num_tokens": 604232273.0, "step": 15839 }, { "epoch": 2.01501081287368, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.409979820251465, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8794049024581909, "num_tokens": 604265236.0, "step": 15840 }, { "epoch": 2.0151380231522706, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.36948013305664, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8872838020324707, "num_tokens": 604304784.0, "step": 15841 }, { "epoch": 2.015265233430861, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.445005416870117, "learning_rate": 1e-06, "loss": 0.3197, "mean_token_accuracy": 0.8984827995300293, "num_tokens": 604341461.0, "step": 15842 }, { "epoch": 2.0153924437094517, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.331905364990234, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8796108365058899, "num_tokens": 604378628.0, "step": 15843 }, { "epoch": 2.0155196539880422, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.3458890914917, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8835572004318237, "num_tokens": 604418070.0, "step": 15844 }, { "epoch": 2.0156468642666328, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.41666316986084, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8862031698226929, "num_tokens": 604454445.0, "step": 15845 }, { "epoch": 2.0157740745452233, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011110305786132812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.381053924560547, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8775566220283508, "num_tokens": 604488559.0, "step": 15846 }, { "epoch": 2.015901284823814, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.39699649810791, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8663842678070068, "num_tokens": 604522538.0, "step": 15847 }, { "epoch": 2.0160284951024043, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.316246032714844, "learning_rate": 1e-06, "loss": 0.3285, "mean_token_accuracy": 0.8921514749526978, "num_tokens": 604561750.0, "step": 15848 }, { "epoch": 2.016155705380995, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.485518455505371, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8865585327148438, "num_tokens": 604594961.0, "step": 15849 }, { "epoch": 2.0162829156595854, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.379668235778809, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8654286861419678, "num_tokens": 604631301.0, "step": 15850 }, { "epoch": 2.016410125938176, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.308777809143066, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8541406989097595, "num_tokens": 604666049.0, "step": 15851 }, { "epoch": 2.0165373362167665, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.397466659545898, "learning_rate": 1e-06, "loss": 0.3327, "mean_token_accuracy": 0.8947707414627075, "num_tokens": 604701984.0, "step": 15852 }, { "epoch": 2.016664546495357, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.393163681030273, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8832536935806274, "num_tokens": 604738414.0, "step": 15853 }, { "epoch": 2.0167917567739475, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.399818420410156, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8764496445655823, "num_tokens": 604779961.0, "step": 15854 }, { "epoch": 2.016918967052538, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.43358325958252, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8676445484161377, "num_tokens": 604819729.0, "step": 15855 }, { "epoch": 2.0170461773311286, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.396797180175781, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8840295076370239, "num_tokens": 604864108.0, "step": 15856 }, { "epoch": 2.0171733876097186, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.42827320098877, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8687588572502136, "num_tokens": 604905582.0, "step": 15857 }, { "epoch": 2.017300597888309, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.439998626708984, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8818502426147461, "num_tokens": 604943931.0, "step": 15858 }, { "epoch": 2.0174278081668997, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.387992858886719, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.869803249835968, "num_tokens": 604982100.0, "step": 15859 }, { "epoch": 2.0175550184454902, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.405492782592773, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8631876111030579, "num_tokens": 605026289.0, "step": 15860 }, { "epoch": 2.0176822287240808, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.369922637939453, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8782873749732971, "num_tokens": 605070214.0, "step": 15861 }, { "epoch": 2.0178094390026713, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.40418815612793, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8699808120727539, "num_tokens": 605113495.0, "step": 15862 }, { "epoch": 2.017936649281262, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.445731163024902, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8782090544700623, "num_tokens": 605145717.0, "step": 15863 }, { "epoch": 2.0180638595598523, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.476119995117188, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8669273257255554, "num_tokens": 605180930.0, "step": 15864 }, { "epoch": 2.018191069838443, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.434126853942871, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8809399008750916, "num_tokens": 605214004.0, "step": 15865 }, { "epoch": 2.0183182801170334, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.407613754272461, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8743504285812378, "num_tokens": 605253244.0, "step": 15866 }, { "epoch": 2.018445490395624, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.438328742980957, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8690388202667236, "num_tokens": 605293832.0, "step": 15867 }, { "epoch": 2.0185727006742145, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.355181694030762, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8848673105239868, "num_tokens": 605333811.0, "step": 15868 }, { "epoch": 2.018699910952805, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.343632698059082, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8749006390571594, "num_tokens": 605371560.0, "step": 15869 }, { "epoch": 2.0188271212313955, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.288498878479004, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8713433146476746, "num_tokens": 605410555.0, "step": 15870 }, { "epoch": 2.018954331509986, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.390785217285156, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8682702779769897, "num_tokens": 605451851.0, "step": 15871 }, { "epoch": 2.0190815417885766, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.299263000488281, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8659676313400269, "num_tokens": 605490442.0, "step": 15872 }, { "epoch": 2.019208752067167, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.488459587097168, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8724226951599121, "num_tokens": 605525237.0, "step": 15873 }, { "epoch": 2.0193359623457576, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.406575202941895, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.878694474697113, "num_tokens": 605563624.0, "step": 15874 }, { "epoch": 2.019463172624348, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.363216400146484, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8837474584579468, "num_tokens": 605597419.0, "step": 15875 }, { "epoch": 2.0195903829029387, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.378125190734863, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8667140603065491, "num_tokens": 605639954.0, "step": 15876 }, { "epoch": 2.019717593181529, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.359463691711426, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8749250173568726, "num_tokens": 605678226.0, "step": 15877 }, { "epoch": 2.0198448034601197, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.437856674194336, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8827095031738281, "num_tokens": 605719925.0, "step": 15878 }, { "epoch": 2.0199720137387103, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.337933540344238, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8550108075141907, "num_tokens": 605761282.0, "step": 15879 }, { "epoch": 2.020099224017301, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.459329605102539, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8831928968429565, "num_tokens": 605800681.0, "step": 15880 }, { "epoch": 2.0202264342958913, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.415627479553223, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8565847277641296, "num_tokens": 605836282.0, "step": 15881 }, { "epoch": 2.0203536445744814, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.363889694213867, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8759967088699341, "num_tokens": 605880360.0, "step": 15882 }, { "epoch": 2.020480854853072, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.369664192199707, "learning_rate": 1e-06, "loss": 0.3519, "mean_token_accuracy": 0.8839091658592224, "num_tokens": 605918363.0, "step": 15883 }, { "epoch": 2.0206080651316625, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.357818603515625, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.887427806854248, "num_tokens": 605953032.0, "step": 15884 }, { "epoch": 2.020735275410253, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.343435287475586, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8807356357574463, "num_tokens": 605992633.0, "step": 15885 }, { "epoch": 2.0208624856888435, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.38259220123291, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8762466907501221, "num_tokens": 606031776.0, "step": 15886 }, { "epoch": 2.020989695967434, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.381670951843262, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8678569197654724, "num_tokens": 606070614.0, "step": 15887 }, { "epoch": 2.0211169062460246, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.358576774597168, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8731405735015869, "num_tokens": 606115898.0, "step": 15888 }, { "epoch": 2.021244116524615, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.371286392211914, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8784070014953613, "num_tokens": 606150231.0, "step": 15889 }, { "epoch": 2.0213713268032056, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.413740158081055, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8661776781082153, "num_tokens": 606191527.0, "step": 15890 }, { "epoch": 2.021498537081796, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.513412475585938, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8665516972541809, "num_tokens": 606229070.0, "step": 15891 }, { "epoch": 2.0216257473603867, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.379419326782227, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8756939768791199, "num_tokens": 606269555.0, "step": 15892 }, { "epoch": 2.021752957638977, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.460000038146973, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.869236946105957, "num_tokens": 606302040.0, "step": 15893 }, { "epoch": 2.0218801679175677, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.426117897033691, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8831582069396973, "num_tokens": 606341769.0, "step": 15894 }, { "epoch": 2.0220073781961583, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.38905143737793, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8612474203109741, "num_tokens": 606379989.0, "step": 15895 }, { "epoch": 2.022134588474749, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.398343086242676, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8692632913589478, "num_tokens": 606414396.0, "step": 15896 }, { "epoch": 2.0222617987533393, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.451128959655762, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.877942681312561, "num_tokens": 606445949.0, "step": 15897 }, { "epoch": 2.02238900903193, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.387922286987305, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8763443231582642, "num_tokens": 606484751.0, "step": 15898 }, { "epoch": 2.0225162193105204, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.387250900268555, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8769190311431885, "num_tokens": 606525390.0, "step": 15899 }, { "epoch": 2.022643429589111, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.439282417297363, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8707938194274902, "num_tokens": 606563364.0, "step": 15900 }, { "epoch": 2.0227706398677014, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.389053344726562, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8764696717262268, "num_tokens": 606598140.0, "step": 15901 }, { "epoch": 2.022897850146292, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.358466148376465, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8665869235992432, "num_tokens": 606636480.0, "step": 15902 }, { "epoch": 2.0230250604248825, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.4719820022583, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8778413534164429, "num_tokens": 606676044.0, "step": 15903 }, { "epoch": 2.023152270703473, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.360844612121582, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8676191568374634, "num_tokens": 606713262.0, "step": 15904 }, { "epoch": 2.0232794809820636, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.395492553710938, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8711796402931213, "num_tokens": 606754844.0, "step": 15905 }, { "epoch": 2.0234066912606536, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.374553680419922, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8739304542541504, "num_tokens": 606800910.0, "step": 15906 }, { "epoch": 2.023533901539244, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.467061042785645, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8692173957824707, "num_tokens": 606841855.0, "step": 15907 }, { "epoch": 2.0236611118178347, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.377701759338379, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8635872006416321, "num_tokens": 606888261.0, "step": 15908 }, { "epoch": 2.0237883220964252, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.434786796569824, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8650215864181519, "num_tokens": 606926235.0, "step": 15909 }, { "epoch": 2.0239155323750158, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.425371170043945, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8694561719894409, "num_tokens": 606962527.0, "step": 15910 }, { "epoch": 2.0240427426536063, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.473899841308594, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8787344694137573, "num_tokens": 607008924.0, "step": 15911 }, { "epoch": 2.024169952932197, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.365802764892578, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8703494071960449, "num_tokens": 607049788.0, "step": 15912 }, { "epoch": 2.0242971632107873, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.419466972351074, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8667685985565186, "num_tokens": 607087592.0, "step": 15913 }, { "epoch": 2.024424373489378, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.4364652633667, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.8886754512786865, "num_tokens": 607119954.0, "step": 15914 }, { "epoch": 2.0245515837679684, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.440041542053223, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8639664649963379, "num_tokens": 607156511.0, "step": 15915 }, { "epoch": 2.024678794046559, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.44511890411377, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8836199045181274, "num_tokens": 607198292.0, "step": 15916 }, { "epoch": 2.0248060043251495, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.493401527404785, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8713490962982178, "num_tokens": 607237572.0, "step": 15917 }, { "epoch": 2.02493321460374, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.412315368652344, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.864781379699707, "num_tokens": 607282321.0, "step": 15918 }, { "epoch": 2.0250604248823305, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.412332534790039, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8719234466552734, "num_tokens": 607313355.0, "step": 15919 }, { "epoch": 2.025187635160921, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.454802513122559, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8713603615760803, "num_tokens": 607351961.0, "step": 15920 }, { "epoch": 2.0253148454395116, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.376457214355469, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8760806918144226, "num_tokens": 607391368.0, "step": 15921 }, { "epoch": 2.025442055718102, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.39627456665039, "learning_rate": 1e-06, "loss": 0.3355, "mean_token_accuracy": 0.8920010328292847, "num_tokens": 607427328.0, "step": 15922 }, { "epoch": 2.0255692659966926, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.415095329284668, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8698525428771973, "num_tokens": 607468243.0, "step": 15923 }, { "epoch": 2.025696476275283, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.486321449279785, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8579021096229553, "num_tokens": 607502466.0, "step": 15924 }, { "epoch": 2.0258236865538737, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.405107498168945, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8644536733627319, "num_tokens": 607544093.0, "step": 15925 }, { "epoch": 2.025950896832464, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.404438972473145, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8800819516181946, "num_tokens": 607576054.0, "step": 15926 }, { "epoch": 2.0260781071110547, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.476670265197754, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8863998651504517, "num_tokens": 607608462.0, "step": 15927 }, { "epoch": 2.0262053173896453, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.374110221862793, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8673017024993896, "num_tokens": 607645272.0, "step": 15928 }, { "epoch": 2.026332527668236, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.448482513427734, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.876349687576294, "num_tokens": 607682585.0, "step": 15929 }, { "epoch": 2.0264597379468263, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.414314270019531, "learning_rate": 1e-06, "loss": 0.3411, "mean_token_accuracy": 0.8906989097595215, "num_tokens": 607721302.0, "step": 15930 }, { "epoch": 2.0265869482254164, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.385567665100098, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8697479963302612, "num_tokens": 607758357.0, "step": 15931 }, { "epoch": 2.026714158504007, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.397547721862793, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.873855471611023, "num_tokens": 607795467.0, "step": 15932 }, { "epoch": 2.0268413687825975, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.454002380371094, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8684971332550049, "num_tokens": 607833192.0, "step": 15933 }, { "epoch": 2.026968579061188, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.440851211547852, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8826063275337219, "num_tokens": 607866672.0, "step": 15934 }, { "epoch": 2.0270957893397785, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.452377319335938, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8748630285263062, "num_tokens": 607907976.0, "step": 15935 }, { "epoch": 2.027222999618369, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.342669486999512, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8770769238471985, "num_tokens": 607946009.0, "step": 15936 }, { "epoch": 2.0273502098969596, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.349475860595703, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.879128098487854, "num_tokens": 607986800.0, "step": 15937 }, { "epoch": 2.02747742017555, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.366640090942383, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8687443733215332, "num_tokens": 608021807.0, "step": 15938 }, { "epoch": 2.0276046304541406, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.429259300231934, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8722173571586609, "num_tokens": 608060294.0, "step": 15939 }, { "epoch": 2.027731840732731, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.298321723937988, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8730728626251221, "num_tokens": 608100790.0, "step": 15940 }, { "epoch": 2.0278590510113217, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.417937278747559, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8701248168945312, "num_tokens": 608142164.0, "step": 15941 }, { "epoch": 2.027986261289912, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.35538101196289, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8751966953277588, "num_tokens": 608184175.0, "step": 15942 }, { "epoch": 2.0281134715685027, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.415853500366211, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8686225414276123, "num_tokens": 608226801.0, "step": 15943 }, { "epoch": 2.0282406818470933, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.435992240905762, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8815006017684937, "num_tokens": 608265155.0, "step": 15944 }, { "epoch": 2.028367892125684, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.375693321228027, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8843874931335449, "num_tokens": 608303817.0, "step": 15945 }, { "epoch": 2.0284951024042743, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.415858268737793, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.880373477935791, "num_tokens": 608338845.0, "step": 15946 }, { "epoch": 2.028622312682865, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.427858352661133, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8736233711242676, "num_tokens": 608377661.0, "step": 15947 }, { "epoch": 2.0287495229614554, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.379871368408203, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8736674785614014, "num_tokens": 608412897.0, "step": 15948 }, { "epoch": 2.028876733240046, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.413657188415527, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8651376366615295, "num_tokens": 608453584.0, "step": 15949 }, { "epoch": 2.0290039435186364, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.42268180847168, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8706709146499634, "num_tokens": 608492211.0, "step": 15950 }, { "epoch": 2.029131153797227, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.315315246582031, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.870601236820221, "num_tokens": 608534317.0, "step": 15951 }, { "epoch": 2.0292583640758175, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.44781494140625, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8830187916755676, "num_tokens": 608569763.0, "step": 15952 }, { "epoch": 2.029385574354408, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.36026668548584, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8770583271980286, "num_tokens": 608612145.0, "step": 15953 }, { "epoch": 2.0295127846329986, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.417914390563965, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8737781047821045, "num_tokens": 608649528.0, "step": 15954 }, { "epoch": 2.0296399949115886, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.459956169128418, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8669340014457703, "num_tokens": 608686973.0, "step": 15955 }, { "epoch": 2.029767205190179, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.352733612060547, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8761798739433289, "num_tokens": 608729169.0, "step": 15956 }, { "epoch": 2.0298944154687697, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.481678009033203, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8730563521385193, "num_tokens": 608764502.0, "step": 15957 }, { "epoch": 2.0300216257473602, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.372981071472168, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8747170567512512, "num_tokens": 608805190.0, "step": 15958 }, { "epoch": 2.0301488360259508, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.46545124053955, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8811118602752686, "num_tokens": 608840006.0, "step": 15959 }, { "epoch": 2.0302760463045413, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.473278045654297, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8815075755119324, "num_tokens": 608871551.0, "step": 15960 }, { "epoch": 2.030403256583132, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.340720176696777, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8525598049163818, "num_tokens": 608909610.0, "step": 15961 }, { "epoch": 2.0305304668617223, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.440159797668457, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8704214096069336, "num_tokens": 608950758.0, "step": 15962 }, { "epoch": 2.030657677140313, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.385041236877441, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8608806133270264, "num_tokens": 608990695.0, "step": 15963 }, { "epoch": 2.0307848874189034, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.39384651184082, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8876007795333862, "num_tokens": 609030978.0, "step": 15964 }, { "epoch": 2.030912097697494, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.366349220275879, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.883441686630249, "num_tokens": 609071974.0, "step": 15965 }, { "epoch": 2.0310393079760845, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.446471214294434, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8761284351348877, "num_tokens": 609111268.0, "step": 15966 }, { "epoch": 2.031166518254675, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.404972076416016, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8692268133163452, "num_tokens": 609149681.0, "step": 15967 }, { "epoch": 2.0312937285332655, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.418153762817383, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8786904811859131, "num_tokens": 609189121.0, "step": 15968 }, { "epoch": 2.031420938811856, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.40639591217041, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.882850170135498, "num_tokens": 609228070.0, "step": 15969 }, { "epoch": 2.0315481490904466, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.415950775146484, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8769078254699707, "num_tokens": 609269718.0, "step": 15970 }, { "epoch": 2.031675359369037, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.388012886047363, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8764241933822632, "num_tokens": 609307236.0, "step": 15971 }, { "epoch": 2.0318025696476276, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.393156051635742, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8569915294647217, "num_tokens": 609342993.0, "step": 15972 }, { "epoch": 2.031929779926218, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.389554023742676, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8800853490829468, "num_tokens": 609376817.0, "step": 15973 }, { "epoch": 2.0320569902048087, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.422686576843262, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8729380965232849, "num_tokens": 609411209.0, "step": 15974 }, { "epoch": 2.032184200483399, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.390096664428711, "learning_rate": 1e-06, "loss": 0.3396, "mean_token_accuracy": 0.8911005258560181, "num_tokens": 609449055.0, "step": 15975 }, { "epoch": 2.0323114107619897, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.423820495605469, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8753656148910522, "num_tokens": 609486332.0, "step": 15976 }, { "epoch": 2.0324386210405803, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.433028221130371, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8612504005432129, "num_tokens": 609526924.0, "step": 15977 }, { "epoch": 2.032565831319171, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.428284645080566, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.887178361415863, "num_tokens": 609559542.0, "step": 15978 }, { "epoch": 2.032693041597761, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.374298095703125, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8670113682746887, "num_tokens": 609598989.0, "step": 15979 }, { "epoch": 2.0328202518763514, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.411933898925781, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.862140953540802, "num_tokens": 609635790.0, "step": 15980 }, { "epoch": 2.032947462154942, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.404581069946289, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8662346601486206, "num_tokens": 609670120.0, "step": 15981 }, { "epoch": 2.0330746724335325, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.39439582824707, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8732609748840332, "num_tokens": 609713260.0, "step": 15982 }, { "epoch": 2.033201882712123, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.437567710876465, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8811681866645813, "num_tokens": 609753236.0, "step": 15983 }, { "epoch": 2.0333290929907135, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.421977043151855, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8775707483291626, "num_tokens": 609791629.0, "step": 15984 }, { "epoch": 2.033456303269304, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.45334529876709, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8584811091423035, "num_tokens": 609833457.0, "step": 15985 }, { "epoch": 2.0335835135478946, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.371005058288574, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8668007254600525, "num_tokens": 609866007.0, "step": 15986 }, { "epoch": 2.033710723826485, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.396035194396973, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.878379225730896, "num_tokens": 609903346.0, "step": 15987 }, { "epoch": 2.0338379341050756, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.391793251037598, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8780235648155212, "num_tokens": 609935885.0, "step": 15988 }, { "epoch": 2.033965144383666, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.420896530151367, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8674803376197815, "num_tokens": 609971127.0, "step": 15989 }, { "epoch": 2.0340923546622567, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.3884916305542, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8650612831115723, "num_tokens": 610016719.0, "step": 15990 }, { "epoch": 2.034219564940847, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.405169486999512, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.86295485496521, "num_tokens": 610057002.0, "step": 15991 }, { "epoch": 2.0343467752194377, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.34310245513916, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8794482350349426, "num_tokens": 610092779.0, "step": 15992 }, { "epoch": 2.0344739854980283, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.407817840576172, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8641767501831055, "num_tokens": 610137092.0, "step": 15993 }, { "epoch": 2.034601195776619, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.402009963989258, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8832941055297852, "num_tokens": 610178632.0, "step": 15994 }, { "epoch": 2.0347284060552093, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.45968246459961, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8726906776428223, "num_tokens": 610219789.0, "step": 15995 }, { "epoch": 2.0348556163338, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.402938842773438, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8599344491958618, "num_tokens": 610259104.0, "step": 15996 }, { "epoch": 2.0349828266123904, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.451916694641113, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8675863146781921, "num_tokens": 610296317.0, "step": 15997 }, { "epoch": 2.035110036890981, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.521190643310547, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8720681071281433, "num_tokens": 610339300.0, "step": 15998 }, { "epoch": 2.0352372471695714, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.359293937683105, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8705582618713379, "num_tokens": 610373917.0, "step": 15999 }, { "epoch": 2.035364457448162, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.514605522155762, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8723962306976318, "num_tokens": 610419571.0, "step": 16000 }, { "epoch": 2.0354916677267525, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.372118949890137, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8649673461914062, "num_tokens": 610458073.0, "step": 16001 }, { "epoch": 2.035618878005343, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.41515064239502, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.863070011138916, "num_tokens": 610497223.0, "step": 16002 }, { "epoch": 2.0357460882839336, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.487149238586426, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8540458083152771, "num_tokens": 610532563.0, "step": 16003 }, { "epoch": 2.0358732985625236, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.34567642211914, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8852517008781433, "num_tokens": 610566644.0, "step": 16004 }, { "epoch": 2.036000508841114, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.421364784240723, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8614913821220398, "num_tokens": 610604755.0, "step": 16005 }, { "epoch": 2.0361277191197047, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.412109375, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8777701258659363, "num_tokens": 610645905.0, "step": 16006 }, { "epoch": 2.036254929398295, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.390615463256836, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.889975368976593, "num_tokens": 610686800.0, "step": 16007 }, { "epoch": 2.0363821396768857, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.406987190246582, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8739084601402283, "num_tokens": 610728291.0, "step": 16008 }, { "epoch": 2.0365093499554763, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.455766677856445, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8718005418777466, "num_tokens": 610759818.0, "step": 16009 }, { "epoch": 2.036636560234067, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.409295082092285, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8596663475036621, "num_tokens": 610800125.0, "step": 16010 }, { "epoch": 2.0367637705126573, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.451643943786621, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8644564151763916, "num_tokens": 610839261.0, "step": 16011 }, { "epoch": 2.036890980791248, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.413168907165527, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8781060576438904, "num_tokens": 610879092.0, "step": 16012 }, { "epoch": 2.0370181910698384, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.449618339538574, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8734816908836365, "num_tokens": 610914494.0, "step": 16013 }, { "epoch": 2.037145401348429, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.361515045166016, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.883314847946167, "num_tokens": 610950784.0, "step": 16014 }, { "epoch": 2.0372726116270194, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.353163719177246, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8640299439430237, "num_tokens": 610989254.0, "step": 16015 }, { "epoch": 2.03739982190561, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.453598022460938, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8658990859985352, "num_tokens": 611022668.0, "step": 16016 }, { "epoch": 2.0375270321842005, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.347890853881836, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8724926114082336, "num_tokens": 611062110.0, "step": 16017 }, { "epoch": 2.037654242462791, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.458551406860352, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8779357671737671, "num_tokens": 611098663.0, "step": 16018 }, { "epoch": 2.0377814527413816, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.409313201904297, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8695432543754578, "num_tokens": 611134058.0, "step": 16019 }, { "epoch": 2.037908663019972, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.430085182189941, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8731556534767151, "num_tokens": 611174748.0, "step": 16020 }, { "epoch": 2.0380358732985626, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.408326148986816, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.8897850513458252, "num_tokens": 611216792.0, "step": 16021 }, { "epoch": 2.038163083577153, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.51453971862793, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8831521272659302, "num_tokens": 611253996.0, "step": 16022 }, { "epoch": 2.0382902938557437, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.368148803710938, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8838213682174683, "num_tokens": 611295351.0, "step": 16023 }, { "epoch": 2.038417504134334, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.4815034866333, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8727161884307861, "num_tokens": 611330513.0, "step": 16024 }, { "epoch": 2.0385447144129247, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.4700288772583, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8817272186279297, "num_tokens": 611369679.0, "step": 16025 }, { "epoch": 2.0386719246915153, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.465065002441406, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8813071846961975, "num_tokens": 611410398.0, "step": 16026 }, { "epoch": 2.038799134970106, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.438469886779785, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8675262928009033, "num_tokens": 611447151.0, "step": 16027 }, { "epoch": 2.0389263452486963, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.475485801696777, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8649550676345825, "num_tokens": 611486923.0, "step": 16028 }, { "epoch": 2.0390535555272864, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.528547286987305, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.879707932472229, "num_tokens": 611519502.0, "step": 16029 }, { "epoch": 2.039180765805877, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.385815620422363, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8602752685546875, "num_tokens": 611557647.0, "step": 16030 }, { "epoch": 2.0393079760844675, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.530180931091309, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8717890977859497, "num_tokens": 611597405.0, "step": 16031 }, { "epoch": 2.039435186363058, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.433181762695312, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8678251504898071, "num_tokens": 611635996.0, "step": 16032 }, { "epoch": 2.0395623966416485, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.49850082397461, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.88178551197052, "num_tokens": 611677900.0, "step": 16033 }, { "epoch": 2.039689606920239, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.323301315307617, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8763336539268494, "num_tokens": 611719184.0, "step": 16034 }, { "epoch": 2.0398168171988296, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.524185180664062, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8722718358039856, "num_tokens": 611761270.0, "step": 16035 }, { "epoch": 2.03994402747742, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.389413833618164, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.873171865940094, "num_tokens": 611794939.0, "step": 16036 }, { "epoch": 2.0400712377560106, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.396480560302734, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8837900161743164, "num_tokens": 611834093.0, "step": 16037 }, { "epoch": 2.040198448034601, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.476759910583496, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8684706687927246, "num_tokens": 611871327.0, "step": 16038 }, { "epoch": 2.0403256583131917, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.473062515258789, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.866662859916687, "num_tokens": 611907837.0, "step": 16039 }, { "epoch": 2.040452868591782, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.434768676757812, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8554865717887878, "num_tokens": 611950964.0, "step": 16040 }, { "epoch": 2.0405800788703727, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.480834007263184, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8692764639854431, "num_tokens": 611985867.0, "step": 16041 }, { "epoch": 2.0407072891489633, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.464299201965332, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8764462471008301, "num_tokens": 612017819.0, "step": 16042 }, { "epoch": 2.040834499427554, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.369174003601074, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.858698844909668, "num_tokens": 612060669.0, "step": 16043 }, { "epoch": 2.0409617097061443, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.518190383911133, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8593533635139465, "num_tokens": 612099801.0, "step": 16044 }, { "epoch": 2.041088919984735, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.381179809570312, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8773421049118042, "num_tokens": 612139454.0, "step": 16045 }, { "epoch": 2.0412161302633254, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.478215217590332, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8546315431594849, "num_tokens": 612181695.0, "step": 16046 }, { "epoch": 2.041343340541916, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.426070213317871, "learning_rate": 1e-06, "loss": 0.353, "mean_token_accuracy": 0.88774573802948, "num_tokens": 612223200.0, "step": 16047 }, { "epoch": 2.0414705508205064, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.41860294342041, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8749605417251587, "num_tokens": 612265946.0, "step": 16048 }, { "epoch": 2.041597761099097, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.407008171081543, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.889137327671051, "num_tokens": 612301246.0, "step": 16049 }, { "epoch": 2.0417249713776875, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.473281860351562, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8584372997283936, "num_tokens": 612337789.0, "step": 16050 }, { "epoch": 2.041852181656278, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.46135425567627, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8858475089073181, "num_tokens": 612372391.0, "step": 16051 }, { "epoch": 2.0419793919348685, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.4465970993042, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8782427310943604, "num_tokens": 612408806.0, "step": 16052 }, { "epoch": 2.0421066022134586, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.416900634765625, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8840124607086182, "num_tokens": 612446349.0, "step": 16053 }, { "epoch": 2.042233812492049, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.453876495361328, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8857139348983765, "num_tokens": 612489899.0, "step": 16054 }, { "epoch": 2.0423610227706397, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.393376350402832, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8679391741752625, "num_tokens": 612532314.0, "step": 16055 }, { "epoch": 2.04248823304923, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.427209854125977, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8808026313781738, "num_tokens": 612568033.0, "step": 16056 }, { "epoch": 2.0426154433278207, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.45037841796875, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8819835186004639, "num_tokens": 612605160.0, "step": 16057 }, { "epoch": 2.0427426536064113, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.469442367553711, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8620713949203491, "num_tokens": 612647851.0, "step": 16058 }, { "epoch": 2.042869863885002, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.421104431152344, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8735559582710266, "num_tokens": 612684770.0, "step": 16059 }, { "epoch": 2.0429970741635923, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.470558166503906, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8628135323524475, "num_tokens": 612721689.0, "step": 16060 }, { "epoch": 2.043124284442183, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.404777526855469, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8728571534156799, "num_tokens": 612757256.0, "step": 16061 }, { "epoch": 2.0432514947207734, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.370373725891113, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8803365230560303, "num_tokens": 612797806.0, "step": 16062 }, { "epoch": 2.043378704999364, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.463997840881348, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.8848486542701721, "num_tokens": 612833507.0, "step": 16063 }, { "epoch": 2.0435059152779544, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.364633560180664, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8795982003211975, "num_tokens": 612872920.0, "step": 16064 }, { "epoch": 2.043633125556545, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.440925598144531, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8794252872467041, "num_tokens": 612909534.0, "step": 16065 }, { "epoch": 2.0437603358351355, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.368484497070312, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8706358671188354, "num_tokens": 612950531.0, "step": 16066 }, { "epoch": 2.043887546113726, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.497323036193848, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8752920627593994, "num_tokens": 612984638.0, "step": 16067 }, { "epoch": 2.0440147563923166, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.412067413330078, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8865872621536255, "num_tokens": 613020632.0, "step": 16068 }, { "epoch": 2.044141966670907, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.536101341247559, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8774123191833496, "num_tokens": 613057381.0, "step": 16069 }, { "epoch": 2.0442691769494976, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.41318130493164, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8729788661003113, "num_tokens": 613094287.0, "step": 16070 }, { "epoch": 2.044396387228088, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.443692207336426, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.880826473236084, "num_tokens": 613126315.0, "step": 16071 }, { "epoch": 2.0445235975066787, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.437597274780273, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8706368803977966, "num_tokens": 613160970.0, "step": 16072 }, { "epoch": 2.044650807785269, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.447504043579102, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.875956654548645, "num_tokens": 613204685.0, "step": 16073 }, { "epoch": 2.0447780180638597, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.4104585647583, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8824788331985474, "num_tokens": 613242394.0, "step": 16074 }, { "epoch": 2.0449052283424503, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.405508041381836, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8814871311187744, "num_tokens": 613283517.0, "step": 16075 }, { "epoch": 2.045032438621041, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.51364803314209, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8793751001358032, "num_tokens": 613317237.0, "step": 16076 }, { "epoch": 2.045159648899631, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.409781455993652, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8759124875068665, "num_tokens": 613355741.0, "step": 16077 }, { "epoch": 2.0452868591782214, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.422040939331055, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8650620579719543, "num_tokens": 613392131.0, "step": 16078 }, { "epoch": 2.045414069456812, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.418046951293945, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8649706840515137, "num_tokens": 613433687.0, "step": 16079 }, { "epoch": 2.0455412797354025, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.365675926208496, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8780590891838074, "num_tokens": 613472882.0, "step": 16080 }, { "epoch": 2.045668490013993, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.424516677856445, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8788622617721558, "num_tokens": 613515869.0, "step": 16081 }, { "epoch": 2.0457957002925835, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.478460311889648, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8754127025604248, "num_tokens": 613551835.0, "step": 16082 }, { "epoch": 2.045922910571174, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.393006324768066, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8723716735839844, "num_tokens": 613592277.0, "step": 16083 }, { "epoch": 2.0460501208497646, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.446852684020996, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8742669820785522, "num_tokens": 613625594.0, "step": 16084 }, { "epoch": 2.046177331128355, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.427719116210938, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8765984773635864, "num_tokens": 613659924.0, "step": 16085 }, { "epoch": 2.0463045414069456, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.43998908996582, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.857451319694519, "num_tokens": 613697067.0, "step": 16086 }, { "epoch": 2.046431751685536, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.474684715270996, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8809229731559753, "num_tokens": 613738211.0, "step": 16087 }, { "epoch": 2.0465589619641267, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.424859046936035, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8793903589248657, "num_tokens": 613782590.0, "step": 16088 }, { "epoch": 2.046686172242717, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.384142875671387, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8808386325836182, "num_tokens": 613825770.0, "step": 16089 }, { "epoch": 2.0468133825213077, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.483393669128418, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8796989917755127, "num_tokens": 613864025.0, "step": 16090 }, { "epoch": 2.0469405927998983, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.392725944519043, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8761913180351257, "num_tokens": 613907181.0, "step": 16091 }, { "epoch": 2.047067803078489, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.418509483337402, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8652595281600952, "num_tokens": 613951258.0, "step": 16092 }, { "epoch": 2.0471950133570793, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.479374885559082, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8728799223899841, "num_tokens": 613991221.0, "step": 16093 }, { "epoch": 2.04732222363567, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.381155967712402, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8841211795806885, "num_tokens": 614026862.0, "step": 16094 }, { "epoch": 2.0474494339142604, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.424555778503418, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8651151657104492, "num_tokens": 614067560.0, "step": 16095 }, { "epoch": 2.047576644192851, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.423121452331543, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8779881000518799, "num_tokens": 614106974.0, "step": 16096 }, { "epoch": 2.0477038544714414, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.39111328125, "learning_rate": 1e-06, "loss": 0.3111, "mean_token_accuracy": 0.900921106338501, "num_tokens": 614145641.0, "step": 16097 }, { "epoch": 2.047831064750032, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.43423843383789, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8776869773864746, "num_tokens": 614181476.0, "step": 16098 }, { "epoch": 2.0479582750286225, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.462496757507324, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.868808388710022, "num_tokens": 614219212.0, "step": 16099 }, { "epoch": 2.048085485307213, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.375696182250977, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8463382720947266, "num_tokens": 614264017.0, "step": 16100 }, { "epoch": 2.0482126955858035, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.401504516601562, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.859326958656311, "num_tokens": 614299031.0, "step": 16101 }, { "epoch": 2.0483399058643936, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.428252220153809, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8800727128982544, "num_tokens": 614334387.0, "step": 16102 }, { "epoch": 2.048467116142984, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.365278244018555, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8752437233924866, "num_tokens": 614372438.0, "step": 16103 }, { "epoch": 2.0485943264215747, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.409976959228516, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8708064556121826, "num_tokens": 614409976.0, "step": 16104 }, { "epoch": 2.048721536700165, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.355316162109375, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8831437826156616, "num_tokens": 614448740.0, "step": 16105 }, { "epoch": 2.0488487469787557, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.46195125579834, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8784384727478027, "num_tokens": 614484644.0, "step": 16106 }, { "epoch": 2.0489759572573463, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.43383502960205, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8555699586868286, "num_tokens": 614521210.0, "step": 16107 }, { "epoch": 2.049103167535937, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.448305130004883, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8691275119781494, "num_tokens": 614561783.0, "step": 16108 }, { "epoch": 2.0492303778145273, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.375707626342773, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.874947190284729, "num_tokens": 614598643.0, "step": 16109 }, { "epoch": 2.049357588093118, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.417080879211426, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8742680549621582, "num_tokens": 614633520.0, "step": 16110 }, { "epoch": 2.0494847983717084, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.372961044311523, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8615410923957825, "num_tokens": 614673266.0, "step": 16111 }, { "epoch": 2.049612008650299, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.374236106872559, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8653194904327393, "num_tokens": 614707201.0, "step": 16112 }, { "epoch": 2.0497392189288894, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.375164031982422, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8791284561157227, "num_tokens": 614746542.0, "step": 16113 }, { "epoch": 2.04986642920748, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.389981269836426, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8639305233955383, "num_tokens": 614781804.0, "step": 16114 }, { "epoch": 2.0499936394860705, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.413774490356445, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8740112781524658, "num_tokens": 614817281.0, "step": 16115 }, { "epoch": 2.050120849764661, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.386439323425293, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.865532398223877, "num_tokens": 614853666.0, "step": 16116 }, { "epoch": 2.0502480600432516, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.37007999420166, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8748826384544373, "num_tokens": 614891221.0, "step": 16117 }, { "epoch": 2.050375270321842, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.393298149108887, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8808974027633667, "num_tokens": 614925956.0, "step": 16118 }, { "epoch": 2.0505024806004326, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.499738693237305, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8796353340148926, "num_tokens": 614959934.0, "step": 16119 }, { "epoch": 2.050629690879023, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.434561729431152, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8710330128669739, "num_tokens": 615000579.0, "step": 16120 }, { "epoch": 2.0507569011576137, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.413662910461426, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8677282333374023, "num_tokens": 615037371.0, "step": 16121 }, { "epoch": 2.050884111436204, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.520987510681152, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8646681308746338, "num_tokens": 615075948.0, "step": 16122 }, { "epoch": 2.0510113217147947, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.402348518371582, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8862944841384888, "num_tokens": 615114619.0, "step": 16123 }, { "epoch": 2.0511385319933853, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.395307540893555, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8746491074562073, "num_tokens": 615149260.0, "step": 16124 }, { "epoch": 2.051265742271976, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.450711250305176, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8773483037948608, "num_tokens": 615183284.0, "step": 16125 }, { "epoch": 2.0513929525505663, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.446754455566406, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8796246647834778, "num_tokens": 615220015.0, "step": 16126 }, { "epoch": 2.0515201628291564, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.473798751831055, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8684033155441284, "num_tokens": 615255614.0, "step": 16127 }, { "epoch": 2.051647373107747, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.481178283691406, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.875129759311676, "num_tokens": 615290232.0, "step": 16128 }, { "epoch": 2.0517745833863374, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.478940963745117, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8801029920578003, "num_tokens": 615328204.0, "step": 16129 }, { "epoch": 2.051901793664928, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.470975875854492, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8652888536453247, "num_tokens": 615362846.0, "step": 16130 }, { "epoch": 2.0520290039435185, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.515361785888672, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8734659552574158, "num_tokens": 615405404.0, "step": 16131 }, { "epoch": 2.052156214222109, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.550315856933594, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8631393313407898, "num_tokens": 615439835.0, "step": 16132 }, { "epoch": 2.0522834245006996, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.461424827575684, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8843262195587158, "num_tokens": 615477832.0, "step": 16133 }, { "epoch": 2.05241063477929, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.565452575683594, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8658348321914673, "num_tokens": 615518089.0, "step": 16134 }, { "epoch": 2.0525378450578806, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.370965003967285, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8792396783828735, "num_tokens": 615557286.0, "step": 16135 }, { "epoch": 2.052665055336471, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.433701515197754, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8774595856666565, "num_tokens": 615593284.0, "step": 16136 }, { "epoch": 2.0527922656150617, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.417718887329102, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.869350790977478, "num_tokens": 615633579.0, "step": 16137 }, { "epoch": 2.052919475893652, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.403313636779785, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8784613609313965, "num_tokens": 615680135.0, "step": 16138 }, { "epoch": 2.0530466861722427, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.385379791259766, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8633167743682861, "num_tokens": 615714801.0, "step": 16139 }, { "epoch": 2.0531738964508333, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.46740436553955, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8661052584648132, "num_tokens": 615750538.0, "step": 16140 }, { "epoch": 2.053301106729424, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.449780464172363, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8795915842056274, "num_tokens": 615789465.0, "step": 16141 }, { "epoch": 2.0534283170080143, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.40283489227295, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8607808351516724, "num_tokens": 615832553.0, "step": 16142 }, { "epoch": 2.053555527286605, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.489611625671387, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8701813817024231, "num_tokens": 615868341.0, "step": 16143 }, { "epoch": 2.0536827375651954, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.397781372070312, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.883202850818634, "num_tokens": 615904324.0, "step": 16144 }, { "epoch": 2.053809947843786, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.37881088256836, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8803698420524597, "num_tokens": 615947656.0, "step": 16145 }, { "epoch": 2.0539371581223764, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.459004402160645, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8680989146232605, "num_tokens": 615989050.0, "step": 16146 }, { "epoch": 2.054064368400967, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.434995651245117, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8781768083572388, "num_tokens": 616027638.0, "step": 16147 }, { "epoch": 2.0541915786795575, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.433905601501465, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8823555707931519, "num_tokens": 616066392.0, "step": 16148 }, { "epoch": 2.054318788958148, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.37642765045166, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8835681676864624, "num_tokens": 616108303.0, "step": 16149 }, { "epoch": 2.0544459992367385, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.382895469665527, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.865647554397583, "num_tokens": 616154174.0, "step": 16150 }, { "epoch": 2.0545732095153286, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.460992813110352, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8724696636199951, "num_tokens": 616193954.0, "step": 16151 }, { "epoch": 2.054700419793919, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.4213228225708, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8785126805305481, "num_tokens": 616232771.0, "step": 16152 }, { "epoch": 2.0548276300725097, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.45716381072998, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8841282725334167, "num_tokens": 616265151.0, "step": 16153 }, { "epoch": 2.0549548403511, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.416472434997559, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.868960976600647, "num_tokens": 616303886.0, "step": 16154 }, { "epoch": 2.0550820506296907, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.484212875366211, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.883512020111084, "num_tokens": 616340040.0, "step": 16155 }, { "epoch": 2.0552092609082813, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.401728630065918, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.879478931427002, "num_tokens": 616371329.0, "step": 16156 }, { "epoch": 2.055336471186872, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.499787330627441, "learning_rate": 1e-06, "loss": 0.3542, "mean_token_accuracy": 0.8881205320358276, "num_tokens": 616406383.0, "step": 16157 }, { "epoch": 2.0554636814654623, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.446725845336914, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8615208268165588, "num_tokens": 616445600.0, "step": 16158 }, { "epoch": 2.055590891744053, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.45106315612793, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8664804697036743, "num_tokens": 616485376.0, "step": 16159 }, { "epoch": 2.0557181020226434, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.453044891357422, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8648636341094971, "num_tokens": 616527104.0, "step": 16160 }, { "epoch": 2.055845312301234, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.429553985595703, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8764039278030396, "num_tokens": 616567557.0, "step": 16161 }, { "epoch": 2.0559725225798244, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.486998558044434, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8793637752532959, "num_tokens": 616601612.0, "step": 16162 }, { "epoch": 2.056099732858415, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.48487663269043, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8720150589942932, "num_tokens": 616638019.0, "step": 16163 }, { "epoch": 2.0562269431370055, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.520949363708496, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8717501759529114, "num_tokens": 616675224.0, "step": 16164 }, { "epoch": 2.056354153415596, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.37520980834961, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8717209100723267, "num_tokens": 616713628.0, "step": 16165 }, { "epoch": 2.0564813636941865, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.502029418945312, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8715083599090576, "num_tokens": 616755985.0, "step": 16166 }, { "epoch": 2.056608573972777, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.407828330993652, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8741490840911865, "num_tokens": 616786564.0, "step": 16167 }, { "epoch": 2.0567357842513676, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.464524269104004, "learning_rate": 1e-06, "loss": 0.4808, "mean_token_accuracy": 0.8474540710449219, "num_tokens": 616823285.0, "step": 16168 }, { "epoch": 2.056862994529958, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.444366455078125, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8819735646247864, "num_tokens": 616864465.0, "step": 16169 }, { "epoch": 2.0569902048085487, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.366424560546875, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8649253249168396, "num_tokens": 616906670.0, "step": 16170 }, { "epoch": 2.057117415087139, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.443737030029297, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8724938631057739, "num_tokens": 616947656.0, "step": 16171 }, { "epoch": 2.0572446253657297, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.4176607131958, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.867615282535553, "num_tokens": 616980161.0, "step": 16172 }, { "epoch": 2.0573718356443202, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.397079467773438, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8731026649475098, "num_tokens": 617015281.0, "step": 16173 }, { "epoch": 2.0574990459229108, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.361989974975586, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.874795138835907, "num_tokens": 617054174.0, "step": 16174 }, { "epoch": 2.057626256201501, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.397826194763184, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8730125427246094, "num_tokens": 617095160.0, "step": 16175 }, { "epoch": 2.0577534664800914, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.426403999328613, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8775239586830139, "num_tokens": 617130781.0, "step": 16176 }, { "epoch": 2.057880676758682, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.388044357299805, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8715708255767822, "num_tokens": 617169473.0, "step": 16177 }, { "epoch": 2.0580078870372724, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.457771301269531, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8682000637054443, "num_tokens": 617212208.0, "step": 16178 }, { "epoch": 2.058135097315863, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.398372650146484, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8751581311225891, "num_tokens": 617246055.0, "step": 16179 }, { "epoch": 2.0582623075944535, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.382678985595703, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8690468072891235, "num_tokens": 617285614.0, "step": 16180 }, { "epoch": 2.058389517873044, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.411904335021973, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8819326162338257, "num_tokens": 617323401.0, "step": 16181 }, { "epoch": 2.0585167281516346, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.414474487304688, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8598132729530334, "num_tokens": 617372550.0, "step": 16182 }, { "epoch": 2.058643938430225, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.403458595275879, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8805351257324219, "num_tokens": 617407894.0, "step": 16183 }, { "epoch": 2.0587711487088156, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.523543357849121, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8654272556304932, "num_tokens": 617450117.0, "step": 16184 }, { "epoch": 2.058898358987406, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.376849174499512, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8791888952255249, "num_tokens": 617485885.0, "step": 16185 }, { "epoch": 2.0590255692659967, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.42812728881836, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8582189083099365, "num_tokens": 617520511.0, "step": 16186 }, { "epoch": 2.059152779544587, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.471930503845215, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.867358922958374, "num_tokens": 617558288.0, "step": 16187 }, { "epoch": 2.0592799898231777, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.356093406677246, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8720167875289917, "num_tokens": 617602991.0, "step": 16188 }, { "epoch": 2.0594072001017683, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.411903381347656, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.871112048625946, "num_tokens": 617639629.0, "step": 16189 }, { "epoch": 2.059534410380359, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.401140213012695, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.867162823677063, "num_tokens": 617676218.0, "step": 16190 }, { "epoch": 2.0596616206589493, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.446114540100098, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8663167953491211, "num_tokens": 617716270.0, "step": 16191 }, { "epoch": 2.05978883093754, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.410890579223633, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8828598856925964, "num_tokens": 617755559.0, "step": 16192 }, { "epoch": 2.0599160412161304, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.42673397064209, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8747463822364807, "num_tokens": 617793522.0, "step": 16193 }, { "epoch": 2.060043251494721, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.418867111206055, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8777062892913818, "num_tokens": 617828005.0, "step": 16194 }, { "epoch": 2.0601704617733114, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.412952423095703, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8695164918899536, "num_tokens": 617863824.0, "step": 16195 }, { "epoch": 2.060297672051902, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.42869758605957, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8686685562133789, "num_tokens": 617903648.0, "step": 16196 }, { "epoch": 2.0604248823304925, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.34685230255127, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8703894019126892, "num_tokens": 617946229.0, "step": 16197 }, { "epoch": 2.060552092609083, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.400986671447754, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8734005093574524, "num_tokens": 617990486.0, "step": 16198 }, { "epoch": 2.0606793028876735, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.432928085327148, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8851866126060486, "num_tokens": 618031743.0, "step": 16199 }, { "epoch": 2.0608065131662636, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.450571060180664, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8710519075393677, "num_tokens": 618070181.0, "step": 16200 }, { "epoch": 2.060933723444854, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.436285018920898, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8796757459640503, "num_tokens": 618106396.0, "step": 16201 }, { "epoch": 2.0610609337234447, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.397773742675781, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8782597184181213, "num_tokens": 618148281.0, "step": 16202 }, { "epoch": 2.061188144002035, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.402645111083984, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8676508069038391, "num_tokens": 618182721.0, "step": 16203 }, { "epoch": 2.0613153542806257, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.49073600769043, "learning_rate": 1e-06, "loss": 0.3509, "mean_token_accuracy": 0.8884605169296265, "num_tokens": 618219497.0, "step": 16204 }, { "epoch": 2.0614425645592163, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.435168266296387, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.890284538269043, "num_tokens": 618258643.0, "step": 16205 }, { "epoch": 2.061569774837807, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.464170455932617, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.8881140947341919, "num_tokens": 618292682.0, "step": 16206 }, { "epoch": 2.0616969851163973, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.423261642456055, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8591576218605042, "num_tokens": 618331375.0, "step": 16207 }, { "epoch": 2.061824195394988, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.38546371459961, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8783796429634094, "num_tokens": 618367452.0, "step": 16208 }, { "epoch": 2.0619514056735784, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.446769714355469, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8724570870399475, "num_tokens": 618410470.0, "step": 16209 }, { "epoch": 2.062078615952169, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.398232460021973, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8661981821060181, "num_tokens": 618447524.0, "step": 16210 }, { "epoch": 2.0622058262307594, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.422443389892578, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8544461131095886, "num_tokens": 618481793.0, "step": 16211 }, { "epoch": 2.06233303650935, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.444639205932617, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8801484107971191, "num_tokens": 618517097.0, "step": 16212 }, { "epoch": 2.0624602467879405, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.373868942260742, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.87143874168396, "num_tokens": 618554221.0, "step": 16213 }, { "epoch": 2.062587457066531, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.449199676513672, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8835500478744507, "num_tokens": 618591279.0, "step": 16214 }, { "epoch": 2.0627146673451215, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.428449630737305, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8611528873443604, "num_tokens": 618628168.0, "step": 16215 }, { "epoch": 2.062841877623712, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.432777404785156, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.878555178642273, "num_tokens": 618664580.0, "step": 16216 }, { "epoch": 2.0629690879023026, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.42102336883545, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8626158833503723, "num_tokens": 618699007.0, "step": 16217 }, { "epoch": 2.063096298180893, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.444344520568848, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8743339776992798, "num_tokens": 618736630.0, "step": 16218 }, { "epoch": 2.0632235084594837, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.46533489227295, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.880628764629364, "num_tokens": 618777868.0, "step": 16219 }, { "epoch": 2.063350718738074, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.480926513671875, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8776915073394775, "num_tokens": 618819134.0, "step": 16220 }, { "epoch": 2.0634779290166647, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.431659698486328, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8739386796951294, "num_tokens": 618862231.0, "step": 16221 }, { "epoch": 2.0636051392952552, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.458963394165039, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8664271831512451, "num_tokens": 618906816.0, "step": 16222 }, { "epoch": 2.0637323495738458, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.434438705444336, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8744723796844482, "num_tokens": 618943239.0, "step": 16223 }, { "epoch": 2.0638595598524363, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.38646411895752, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8870532512664795, "num_tokens": 618978641.0, "step": 16224 }, { "epoch": 2.0639867701310264, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.474287986755371, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8744264841079712, "num_tokens": 619020522.0, "step": 16225 }, { "epoch": 2.064113980409617, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.450010299682617, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8857163786888123, "num_tokens": 619060769.0, "step": 16226 }, { "epoch": 2.0642411906882074, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.512348175048828, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8843634128570557, "num_tokens": 619098711.0, "step": 16227 }, { "epoch": 2.064368400966798, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.434172630310059, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8697249889373779, "num_tokens": 619133673.0, "step": 16228 }, { "epoch": 2.0644956112453885, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.508830070495605, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8679984211921692, "num_tokens": 619174554.0, "step": 16229 }, { "epoch": 2.064622821523979, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.379874229431152, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8630894422531128, "num_tokens": 619211095.0, "step": 16230 }, { "epoch": 2.0647500318025696, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.49746036529541, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.885054349899292, "num_tokens": 619246346.0, "step": 16231 }, { "epoch": 2.06487724208116, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.479286193847656, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8709993958473206, "num_tokens": 619280629.0, "step": 16232 }, { "epoch": 2.0650044523597506, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.543261528015137, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8539035320281982, "num_tokens": 619320928.0, "step": 16233 }, { "epoch": 2.065131662638341, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.398265838623047, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8751212358474731, "num_tokens": 619369747.0, "step": 16234 }, { "epoch": 2.0652588729169317, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.516223907470703, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8759781122207642, "num_tokens": 619408694.0, "step": 16235 }, { "epoch": 2.065386083195522, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.386323928833008, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8814764022827148, "num_tokens": 619446858.0, "step": 16236 }, { "epoch": 2.0655132934741127, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.451441764831543, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8744323253631592, "num_tokens": 619482281.0, "step": 16237 }, { "epoch": 2.0656405037527032, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.449650764465332, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8825364112854004, "num_tokens": 619520309.0, "step": 16238 }, { "epoch": 2.065767714031294, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.473968505859375, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8772178888320923, "num_tokens": 619554662.0, "step": 16239 }, { "epoch": 2.0658949243098843, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.413747787475586, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8713963031768799, "num_tokens": 619599632.0, "step": 16240 }, { "epoch": 2.066022134588475, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.480170249938965, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8859882354736328, "num_tokens": 619639414.0, "step": 16241 }, { "epoch": 2.0661493448670654, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.505243301391602, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8764260411262512, "num_tokens": 619675878.0, "step": 16242 }, { "epoch": 2.066276555145656, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.554898262023926, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8770950436592102, "num_tokens": 619709486.0, "step": 16243 }, { "epoch": 2.0664037654242464, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.41296100616455, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8742191791534424, "num_tokens": 619753576.0, "step": 16244 }, { "epoch": 2.066530975702837, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.557122230529785, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8749584555625916, "num_tokens": 619792520.0, "step": 16245 }, { "epoch": 2.0666581859814275, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.543042182922363, "learning_rate": 1e-06, "loss": 0.4815, "mean_token_accuracy": 0.8534864187240601, "num_tokens": 619827871.0, "step": 16246 }, { "epoch": 2.066785396260018, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.535815238952637, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8695350289344788, "num_tokens": 619864527.0, "step": 16247 }, { "epoch": 2.0669126065386085, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.505813598632812, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8750965595245361, "num_tokens": 619904934.0, "step": 16248 }, { "epoch": 2.0670398168171986, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.48373794555664, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8824109435081482, "num_tokens": 619945472.0, "step": 16249 }, { "epoch": 2.067167027095789, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.493264198303223, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8850176930427551, "num_tokens": 619981738.0, "step": 16250 }, { "epoch": 2.0672942373743797, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.461216926574707, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8760905265808105, "num_tokens": 620024593.0, "step": 16251 }, { "epoch": 2.06742144765297, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.54924488067627, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.880858302116394, "num_tokens": 620064059.0, "step": 16252 }, { "epoch": 2.0675486579315607, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.468185424804688, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8749825358390808, "num_tokens": 620100028.0, "step": 16253 }, { "epoch": 2.0676758682101513, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.523869514465332, "learning_rate": 1e-06, "loss": 0.353, "mean_token_accuracy": 0.8874306678771973, "num_tokens": 620135181.0, "step": 16254 }, { "epoch": 2.067803078488742, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.499175071716309, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8759390115737915, "num_tokens": 620174887.0, "step": 16255 }, { "epoch": 2.0679302887673323, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.465128898620605, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8745151162147522, "num_tokens": 620212924.0, "step": 16256 }, { "epoch": 2.068057499045923, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.468329429626465, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8658353686332703, "num_tokens": 620251088.0, "step": 16257 }, { "epoch": 2.0681847093245134, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.467555046081543, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.87404465675354, "num_tokens": 620287888.0, "step": 16258 }, { "epoch": 2.068311919603104, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.496818542480469, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8721485137939453, "num_tokens": 620324937.0, "step": 16259 }, { "epoch": 2.0684391298816944, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.456233024597168, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8751935958862305, "num_tokens": 620363390.0, "step": 16260 }, { "epoch": 2.068566340160285, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.456382751464844, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8820337057113647, "num_tokens": 620403707.0, "step": 16261 }, { "epoch": 2.0686935504388755, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.499220848083496, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8781743049621582, "num_tokens": 620445994.0, "step": 16262 }, { "epoch": 2.068820760717466, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.440394401550293, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8815963864326477, "num_tokens": 620480883.0, "step": 16263 }, { "epoch": 2.0689479709960565, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.530891418457031, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8874818086624146, "num_tokens": 620518935.0, "step": 16264 }, { "epoch": 2.069075181274647, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.490641593933105, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8688595294952393, "num_tokens": 620551985.0, "step": 16265 }, { "epoch": 2.0692023915532376, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.545263290405273, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8700898885726929, "num_tokens": 620592809.0, "step": 16266 }, { "epoch": 2.069329601831828, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.479637145996094, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8608399033546448, "num_tokens": 620627951.0, "step": 16267 }, { "epoch": 2.0694568121104187, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.490340232849121, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8797026872634888, "num_tokens": 620660892.0, "step": 16268 }, { "epoch": 2.069584022389009, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.463337898254395, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8509693741798401, "num_tokens": 620698165.0, "step": 16269 }, { "epoch": 2.0697112326675997, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.519454956054688, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8660237789154053, "num_tokens": 620734758.0, "step": 16270 }, { "epoch": 2.0698384429461902, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.458600997924805, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8558699488639832, "num_tokens": 620770224.0, "step": 16271 }, { "epoch": 2.0699656532247808, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.536029815673828, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8816279768943787, "num_tokens": 620805724.0, "step": 16272 }, { "epoch": 2.070092863503371, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.429363250732422, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8629389405250549, "num_tokens": 620847622.0, "step": 16273 }, { "epoch": 2.0702200737819614, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.558248519897461, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8711044788360596, "num_tokens": 620879042.0, "step": 16274 }, { "epoch": 2.070347284060552, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.414106369018555, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8703051209449768, "num_tokens": 620922685.0, "step": 16275 }, { "epoch": 2.0704744943391424, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.511094093322754, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8685711622238159, "num_tokens": 620954854.0, "step": 16276 }, { "epoch": 2.070601704617733, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.449740409851074, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8724483847618103, "num_tokens": 620994581.0, "step": 16277 }, { "epoch": 2.0707289148963235, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.484756469726562, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.875160276889801, "num_tokens": 621030052.0, "step": 16278 }, { "epoch": 2.070856125174914, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.44112491607666, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8829489946365356, "num_tokens": 621072474.0, "step": 16279 }, { "epoch": 2.0709833354535045, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.503544807434082, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8774497509002686, "num_tokens": 621104365.0, "step": 16280 }, { "epoch": 2.071110545732095, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.499798774719238, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8792144656181335, "num_tokens": 621140497.0, "step": 16281 }, { "epoch": 2.0712377560106856, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.443601608276367, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.877152681350708, "num_tokens": 621177431.0, "step": 16282 }, { "epoch": 2.071364966289276, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.498903274536133, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8768891096115112, "num_tokens": 621212874.0, "step": 16283 }, { "epoch": 2.0714921765678667, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.432251930236816, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8786039352416992, "num_tokens": 621249927.0, "step": 16284 }, { "epoch": 2.071619386846457, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.546278953552246, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.850619912147522, "num_tokens": 621280794.0, "step": 16285 }, { "epoch": 2.0717465971250477, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.511178970336914, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8585749864578247, "num_tokens": 621321167.0, "step": 16286 }, { "epoch": 2.0718738074036382, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.385573387145996, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8852622509002686, "num_tokens": 621357962.0, "step": 16287 }, { "epoch": 2.0720010176822288, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.475364685058594, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8916306495666504, "num_tokens": 621398933.0, "step": 16288 }, { "epoch": 2.0721282279608193, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.450882911682129, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.8932880163192749, "num_tokens": 621442334.0, "step": 16289 }, { "epoch": 2.07225543823941, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.507598876953125, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8771072626113892, "num_tokens": 621478706.0, "step": 16290 }, { "epoch": 2.0723826485180004, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.480350494384766, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8804436922073364, "num_tokens": 621513911.0, "step": 16291 }, { "epoch": 2.072509858796591, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.436734199523926, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8695152997970581, "num_tokens": 621554080.0, "step": 16292 }, { "epoch": 2.0726370690751814, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.458925247192383, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8765076398849487, "num_tokens": 621592059.0, "step": 16293 }, { "epoch": 2.072764279353772, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.464605331420898, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8854067325592041, "num_tokens": 621629626.0, "step": 16294 }, { "epoch": 2.0728914896323625, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.476204872131348, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8789210319519043, "num_tokens": 621671341.0, "step": 16295 }, { "epoch": 2.073018699910953, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.458904266357422, "learning_rate": 1e-06, "loss": 0.3409, "mean_token_accuracy": 0.8884252309799194, "num_tokens": 621708756.0, "step": 16296 }, { "epoch": 2.0731459101895435, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.498671531677246, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8781532049179077, "num_tokens": 621748009.0, "step": 16297 }, { "epoch": 2.0732731204681336, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.4786958694458, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8830878734588623, "num_tokens": 621788695.0, "step": 16298 }, { "epoch": 2.073400330746724, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.544604301452637, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8736162781715393, "num_tokens": 621828099.0, "step": 16299 }, { "epoch": 2.0735275410253147, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.480770111083984, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.857570469379425, "num_tokens": 621867033.0, "step": 16300 }, { "epoch": 2.073654751303905, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.497024536132812, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8645755052566528, "num_tokens": 621907435.0, "step": 16301 }, { "epoch": 2.0737819615824957, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.540212631225586, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8659777641296387, "num_tokens": 621948995.0, "step": 16302 }, { "epoch": 2.0739091718610863, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.521851539611816, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8864204287528992, "num_tokens": 621991109.0, "step": 16303 }, { "epoch": 2.074036382139677, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.476457595825195, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8806010484695435, "num_tokens": 622026585.0, "step": 16304 }, { "epoch": 2.0741635924182673, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.506084442138672, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8842369318008423, "num_tokens": 622065208.0, "step": 16305 }, { "epoch": 2.074290802696858, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.463440895080566, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8653132915496826, "num_tokens": 622104838.0, "step": 16306 }, { "epoch": 2.0744180129754484, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.518502235412598, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8653662204742432, "num_tokens": 622136879.0, "step": 16307 }, { "epoch": 2.074545223254039, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.47828197479248, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.885326087474823, "num_tokens": 622175938.0, "step": 16308 }, { "epoch": 2.0746724335326294, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.458179473876953, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8830646872520447, "num_tokens": 622216663.0, "step": 16309 }, { "epoch": 2.07479964381122, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.447213172912598, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8829671144485474, "num_tokens": 622250450.0, "step": 16310 }, { "epoch": 2.0749268540898105, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.46768856048584, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8662760257720947, "num_tokens": 622291919.0, "step": 16311 }, { "epoch": 2.075054064368401, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.501526832580566, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8734020590782166, "num_tokens": 622325892.0, "step": 16312 }, { "epoch": 2.0751812746469915, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.484566688537598, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8742128610610962, "num_tokens": 622356966.0, "step": 16313 }, { "epoch": 2.075308484925582, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.525181770324707, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.877487301826477, "num_tokens": 622389931.0, "step": 16314 }, { "epoch": 2.0754356952041726, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.447173118591309, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8735643625259399, "num_tokens": 622427817.0, "step": 16315 }, { "epoch": 2.075562905482763, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.481297492980957, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8652433156967163, "num_tokens": 622470353.0, "step": 16316 }, { "epoch": 2.0756901157613536, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.444519996643066, "learning_rate": 1e-06, "loss": 0.3369, "mean_token_accuracy": 0.8897898197174072, "num_tokens": 622507577.0, "step": 16317 }, { "epoch": 2.075817326039944, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.516666412353516, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8700551986694336, "num_tokens": 622544746.0, "step": 16318 }, { "epoch": 2.0759445363185347, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.489474296569824, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8679908514022827, "num_tokens": 622583688.0, "step": 16319 }, { "epoch": 2.0760717465971252, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.542085647583008, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8748586177825928, "num_tokens": 622619911.0, "step": 16320 }, { "epoch": 2.0761989568757158, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.443321228027344, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8805069923400879, "num_tokens": 622665732.0, "step": 16321 }, { "epoch": 2.0763261671543063, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.435503959655762, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8817362785339355, "num_tokens": 622705746.0, "step": 16322 }, { "epoch": 2.0764533774328964, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.459640502929688, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8762802481651306, "num_tokens": 622745481.0, "step": 16323 }, { "epoch": 2.076580587711487, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.41976547241211, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8771803379058838, "num_tokens": 622786910.0, "step": 16324 }, { "epoch": 2.0767077979900774, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.559601783752441, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8702249526977539, "num_tokens": 622826976.0, "step": 16325 }, { "epoch": 2.076835008268668, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.51058578491211, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.875299334526062, "num_tokens": 622858945.0, "step": 16326 }, { "epoch": 2.0769622185472585, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.499640464782715, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8862729072570801, "num_tokens": 622893229.0, "step": 16327 }, { "epoch": 2.077089428825849, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.443958282470703, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8849359154701233, "num_tokens": 622933701.0, "step": 16328 }, { "epoch": 2.0772166391044395, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.53213882446289, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8612948060035706, "num_tokens": 622971226.0, "step": 16329 }, { "epoch": 2.07734384938303, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.46846866607666, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8719625473022461, "num_tokens": 623009232.0, "step": 16330 }, { "epoch": 2.0774710596616206, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.420141220092773, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8690599799156189, "num_tokens": 623048005.0, "step": 16331 }, { "epoch": 2.077598269940211, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.46079158782959, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8745815753936768, "num_tokens": 623088474.0, "step": 16332 }, { "epoch": 2.0777254802188017, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.504616737365723, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8754575252532959, "num_tokens": 623123700.0, "step": 16333 }, { "epoch": 2.077852690497392, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.495770454406738, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8774318099021912, "num_tokens": 623161406.0, "step": 16334 }, { "epoch": 2.0779799007759827, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.492239952087402, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8657174706459045, "num_tokens": 623199770.0, "step": 16335 }, { "epoch": 2.0781071110545732, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.460392951965332, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8782879114151001, "num_tokens": 623234635.0, "step": 16336 }, { "epoch": 2.0782343213331638, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.455082893371582, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.876456081867218, "num_tokens": 623273695.0, "step": 16337 }, { "epoch": 2.0783615316117543, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.475602149963379, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8604125380516052, "num_tokens": 623315278.0, "step": 16338 }, { "epoch": 2.078488741890345, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.501717567443848, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8835023641586304, "num_tokens": 623351753.0, "step": 16339 }, { "epoch": 2.0786159521689354, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.505598068237305, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8574718832969666, "num_tokens": 623387812.0, "step": 16340 }, { "epoch": 2.078743162447526, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.483430862426758, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8782655000686646, "num_tokens": 623426784.0, "step": 16341 }, { "epoch": 2.0788703727261164, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.48133373260498, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.8519088625907898, "num_tokens": 623463362.0, "step": 16342 }, { "epoch": 2.078997583004707, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.468656539916992, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8675061464309692, "num_tokens": 623500108.0, "step": 16343 }, { "epoch": 2.0791247932832975, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.449698448181152, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8709700703620911, "num_tokens": 623537159.0, "step": 16344 }, { "epoch": 2.079252003561888, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.471546173095703, "learning_rate": 1e-06, "loss": 0.3392, "mean_token_accuracy": 0.8930013179779053, "num_tokens": 623576256.0, "step": 16345 }, { "epoch": 2.0793792138404785, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.482643127441406, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8711004257202148, "num_tokens": 623612454.0, "step": 16346 }, { "epoch": 2.0795064241190686, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.493889808654785, "learning_rate": 1e-06, "loss": 0.3477, "mean_token_accuracy": 0.8876370191574097, "num_tokens": 623654392.0, "step": 16347 }, { "epoch": 2.079633634397659, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.428915977478027, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8564149141311646, "num_tokens": 623697350.0, "step": 16348 }, { "epoch": 2.0797608446762497, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.497899055480957, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8781789541244507, "num_tokens": 623736272.0, "step": 16349 }, { "epoch": 2.07988805495484, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.52011775970459, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8651947975158691, "num_tokens": 623781319.0, "step": 16350 }, { "epoch": 2.0800152652334307, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.448195457458496, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8752536773681641, "num_tokens": 623821944.0, "step": 16351 }, { "epoch": 2.0801424755120212, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.447670936584473, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8697432279586792, "num_tokens": 623858399.0, "step": 16352 }, { "epoch": 2.0802696857906118, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.382257461547852, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8756837248802185, "num_tokens": 623897839.0, "step": 16353 }, { "epoch": 2.0803968960692023, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.44918155670166, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8587360978126526, "num_tokens": 623933127.0, "step": 16354 }, { "epoch": 2.080524106347793, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.407980918884277, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8741973638534546, "num_tokens": 623972140.0, "step": 16355 }, { "epoch": 2.0806513166263834, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.476078033447266, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8642324805259705, "num_tokens": 624008875.0, "step": 16356 }, { "epoch": 2.080778526904974, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.424701690673828, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8713864088058472, "num_tokens": 624045162.0, "step": 16357 }, { "epoch": 2.0809057371835644, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.408249855041504, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8832192420959473, "num_tokens": 624085275.0, "step": 16358 }, { "epoch": 2.081032947462155, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.42520523071289, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8719075918197632, "num_tokens": 624127479.0, "step": 16359 }, { "epoch": 2.0811601577407455, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.443130493164062, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.876455545425415, "num_tokens": 624165231.0, "step": 16360 }, { "epoch": 2.081287368019336, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.423781394958496, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8632380366325378, "num_tokens": 624205109.0, "step": 16361 }, { "epoch": 2.0814145782979265, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.397791862487793, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8700078725814819, "num_tokens": 624241849.0, "step": 16362 }, { "epoch": 2.081541788576517, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.456183433532715, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.875988781452179, "num_tokens": 624274174.0, "step": 16363 }, { "epoch": 2.0816689988551076, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.469532012939453, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8725839257240295, "num_tokens": 624311526.0, "step": 16364 }, { "epoch": 2.081796209133698, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.562708854675293, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8604058623313904, "num_tokens": 624346657.0, "step": 16365 }, { "epoch": 2.0819234194122886, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.416703224182129, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8585907816886902, "num_tokens": 624386563.0, "step": 16366 }, { "epoch": 2.082050629690879, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.465682029724121, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8834007978439331, "num_tokens": 624427299.0, "step": 16367 }, { "epoch": 2.0821778399694697, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.455337524414062, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8572306036949158, "num_tokens": 624461253.0, "step": 16368 }, { "epoch": 2.0823050502480602, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.477755546569824, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8830283880233765, "num_tokens": 624498699.0, "step": 16369 }, { "epoch": 2.0824322605266508, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.479243278503418, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8715196847915649, "num_tokens": 624534655.0, "step": 16370 }, { "epoch": 2.082559470805241, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.511621475219727, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8775638341903687, "num_tokens": 624575342.0, "step": 16371 }, { "epoch": 2.0826866810838314, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.527419090270996, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8792653679847717, "num_tokens": 624608493.0, "step": 16372 }, { "epoch": 2.082813891362422, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.516077995300293, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8563757538795471, "num_tokens": 624655777.0, "step": 16373 }, { "epoch": 2.0829411016410124, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.513060569763184, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.883266270160675, "num_tokens": 624699349.0, "step": 16374 }, { "epoch": 2.083068311919603, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.464706420898438, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8817144632339478, "num_tokens": 624730146.0, "step": 16375 }, { "epoch": 2.0831955221981935, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.46068286895752, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8666399717330933, "num_tokens": 624773186.0, "step": 16376 }, { "epoch": 2.083322732476784, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.50636100769043, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8710538148880005, "num_tokens": 624810256.0, "step": 16377 }, { "epoch": 2.0834499427553745, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.555644035339355, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8637997508049011, "num_tokens": 624848166.0, "step": 16378 }, { "epoch": 2.083577153033965, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.426209449768066, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8727748394012451, "num_tokens": 624887601.0, "step": 16379 }, { "epoch": 2.0837043633125556, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.49089241027832, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.873888373374939, "num_tokens": 624921160.0, "step": 16380 }, { "epoch": 2.083831573591146, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.46161937713623, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8665087819099426, "num_tokens": 624959053.0, "step": 16381 }, { "epoch": 2.0839587838697367, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.408270835876465, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8870227932929993, "num_tokens": 624995272.0, "step": 16382 }, { "epoch": 2.084085994148327, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.478964805603027, "learning_rate": 1e-06, "loss": 0.3437, "mean_token_accuracy": 0.8886322379112244, "num_tokens": 625033388.0, "step": 16383 }, { "epoch": 2.0842132044269177, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.419081687927246, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8634421229362488, "num_tokens": 625071103.0, "step": 16384 }, { "epoch": 2.0843404147055082, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.450973510742188, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8856180906295776, "num_tokens": 625109562.0, "step": 16385 }, { "epoch": 2.0844676249840988, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.445048332214355, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8792855739593506, "num_tokens": 625147438.0, "step": 16386 }, { "epoch": 2.0845948352626893, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.451728820800781, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8827728033065796, "num_tokens": 625184293.0, "step": 16387 }, { "epoch": 2.08472204554128, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.478687286376953, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8756552338600159, "num_tokens": 625218201.0, "step": 16388 }, { "epoch": 2.0848492558198704, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.503962516784668, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8720170855522156, "num_tokens": 625252072.0, "step": 16389 }, { "epoch": 2.084976466098461, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.488441467285156, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8858903646469116, "num_tokens": 625286990.0, "step": 16390 }, { "epoch": 2.0851036763770514, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.438542366027832, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8841427564620972, "num_tokens": 625326579.0, "step": 16391 }, { "epoch": 2.085230886655642, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.477051734924316, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8787374496459961, "num_tokens": 625367317.0, "step": 16392 }, { "epoch": 2.0853580969342325, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.465729713439941, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8682953715324402, "num_tokens": 625401047.0, "step": 16393 }, { "epoch": 2.085485307212823, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.487871170043945, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.859147846698761, "num_tokens": 625441878.0, "step": 16394 }, { "epoch": 2.0856125174914135, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.502335548400879, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8722007274627686, "num_tokens": 625477007.0, "step": 16395 }, { "epoch": 2.0857397277700036, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.423982620239258, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8860077261924744, "num_tokens": 625519908.0, "step": 16396 }, { "epoch": 2.085866938048594, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.471423149108887, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8765172362327576, "num_tokens": 625554946.0, "step": 16397 }, { "epoch": 2.0859941483271847, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.478592872619629, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8851300477981567, "num_tokens": 625592121.0, "step": 16398 }, { "epoch": 2.086121358605775, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.422683715820312, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8725416660308838, "num_tokens": 625630596.0, "step": 16399 }, { "epoch": 2.0862485688843657, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.419891357421875, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8562057018280029, "num_tokens": 625669535.0, "step": 16400 }, { "epoch": 2.0863757791629562, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.454242706298828, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8718481063842773, "num_tokens": 625710072.0, "step": 16401 }, { "epoch": 2.0865029894415468, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.430265426635742, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8808619976043701, "num_tokens": 625746328.0, "step": 16402 }, { "epoch": 2.0866301997201373, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.400763511657715, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8747966289520264, "num_tokens": 625787108.0, "step": 16403 }, { "epoch": 2.086757409998728, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.44679069519043, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.877189040184021, "num_tokens": 625827887.0, "step": 16404 }, { "epoch": 2.0868846202773184, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.499866485595703, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8740253448486328, "num_tokens": 625863475.0, "step": 16405 }, { "epoch": 2.087011830555909, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.470524787902832, "learning_rate": 1e-06, "loss": 0.3503, "mean_token_accuracy": 0.8895964026451111, "num_tokens": 625902369.0, "step": 16406 }, { "epoch": 2.0871390408344994, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.446897506713867, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8492905497550964, "num_tokens": 625941054.0, "step": 16407 }, { "epoch": 2.08726625111309, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.461915969848633, "learning_rate": 1e-06, "loss": 0.4672, "mean_token_accuracy": 0.8557653427124023, "num_tokens": 625981113.0, "step": 16408 }, { "epoch": 2.0873934613916805, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.483015060424805, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.870371401309967, "num_tokens": 626014758.0, "step": 16409 }, { "epoch": 2.087520671670271, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.475985527038574, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8597421050071716, "num_tokens": 626054442.0, "step": 16410 }, { "epoch": 2.0876478819488615, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.4227876663208, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8564490079879761, "num_tokens": 626093993.0, "step": 16411 }, { "epoch": 2.087775092227452, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.469483375549316, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8566091060638428, "num_tokens": 626131242.0, "step": 16412 }, { "epoch": 2.0879023025060426, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.48025131225586, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8574220538139343, "num_tokens": 626162137.0, "step": 16413 }, { "epoch": 2.088029512784633, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.46240520477295, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8742393255233765, "num_tokens": 626201031.0, "step": 16414 }, { "epoch": 2.0881567230632236, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.487030029296875, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8830732107162476, "num_tokens": 626243647.0, "step": 16415 }, { "epoch": 2.088283933341814, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.430459022521973, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8791674971580505, "num_tokens": 626281346.0, "step": 16416 }, { "epoch": 2.0884111436204047, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.46516227722168, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8827547430992126, "num_tokens": 626315964.0, "step": 16417 }, { "epoch": 2.0885383538989952, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.388325691223145, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8666650056838989, "num_tokens": 626367920.0, "step": 16418 }, { "epoch": 2.0886655641775858, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.388431549072266, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8700892329216003, "num_tokens": 626412123.0, "step": 16419 }, { "epoch": 2.0887927744561763, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.49842643737793, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8463218212127686, "num_tokens": 626451560.0, "step": 16420 }, { "epoch": 2.0889199847347664, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.41370964050293, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8667150735855103, "num_tokens": 626489361.0, "step": 16421 }, { "epoch": 2.089047195013357, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.445964813232422, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8875962495803833, "num_tokens": 626527150.0, "step": 16422 }, { "epoch": 2.0891744052919474, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.469070434570312, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8667579889297485, "num_tokens": 626568110.0, "step": 16423 }, { "epoch": 2.089301615570538, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.496733665466309, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8744233846664429, "num_tokens": 626607215.0, "step": 16424 }, { "epoch": 2.0894288258491285, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.438597679138184, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8782734274864197, "num_tokens": 626644424.0, "step": 16425 }, { "epoch": 2.089556036127719, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.449599266052246, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8516316413879395, "num_tokens": 626689509.0, "step": 16426 }, { "epoch": 2.0896832464063095, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.436798095703125, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8632641434669495, "num_tokens": 626733866.0, "step": 16427 }, { "epoch": 2.0898104566849, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.498367309570312, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8808636665344238, "num_tokens": 626767465.0, "step": 16428 }, { "epoch": 2.0899376669634906, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.366827011108398, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8595144748687744, "num_tokens": 626806049.0, "step": 16429 }, { "epoch": 2.090064877242081, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.501616477966309, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8598623275756836, "num_tokens": 626848917.0, "step": 16430 }, { "epoch": 2.0901920875206716, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.4682035446167, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8597726821899414, "num_tokens": 626884442.0, "step": 16431 }, { "epoch": 2.090319297799262, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.507612228393555, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8712608814239502, "num_tokens": 626922517.0, "step": 16432 }, { "epoch": 2.0904465080778527, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.467089653015137, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8711673021316528, "num_tokens": 626962880.0, "step": 16433 }, { "epoch": 2.0905737183564432, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.486700057983398, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8799580335617065, "num_tokens": 627008115.0, "step": 16434 }, { "epoch": 2.0907009286350338, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.46143913269043, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8836370706558228, "num_tokens": 627043935.0, "step": 16435 }, { "epoch": 2.0908281389136243, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.48881721496582, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8743849396705627, "num_tokens": 627078495.0, "step": 16436 }, { "epoch": 2.090955349192215, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.477753639221191, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8772265315055847, "num_tokens": 627116652.0, "step": 16437 }, { "epoch": 2.0910825594708053, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.498763084411621, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8808290958404541, "num_tokens": 627152326.0, "step": 16438 }, { "epoch": 2.091209769749396, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.514782905578613, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8822469711303711, "num_tokens": 627189502.0, "step": 16439 }, { "epoch": 2.0913369800279864, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.49129867553711, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.8884163498878479, "num_tokens": 627225438.0, "step": 16440 }, { "epoch": 2.091464190306577, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.540508270263672, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8691362142562866, "num_tokens": 627258154.0, "step": 16441 }, { "epoch": 2.0915914005851675, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.477709770202637, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8649351596832275, "num_tokens": 627301804.0, "step": 16442 }, { "epoch": 2.091718610863758, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.453310012817383, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8767247796058655, "num_tokens": 627344301.0, "step": 16443 }, { "epoch": 2.0918458211423485, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.549324989318848, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8812328577041626, "num_tokens": 627381068.0, "step": 16444 }, { "epoch": 2.0919730314209386, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.478009223937988, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8665831089019775, "num_tokens": 627423410.0, "step": 16445 }, { "epoch": 2.092100241699529, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.498849868774414, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8830981254577637, "num_tokens": 627456593.0, "step": 16446 }, { "epoch": 2.0922274519781197, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.524800300598145, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8758870363235474, "num_tokens": 627493742.0, "step": 16447 }, { "epoch": 2.09235466225671, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.491339683532715, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8804073929786682, "num_tokens": 627529809.0, "step": 16448 }, { "epoch": 2.0924818725353007, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.430381774902344, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8761608004570007, "num_tokens": 627570693.0, "step": 16449 }, { "epoch": 2.0926090828138912, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.478109359741211, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8801844120025635, "num_tokens": 627607608.0, "step": 16450 }, { "epoch": 2.0927362930924818, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.527667045593262, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8711130619049072, "num_tokens": 627642102.0, "step": 16451 }, { "epoch": 2.0928635033710723, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.44149398803711, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8674465417861938, "num_tokens": 627678838.0, "step": 16452 }, { "epoch": 2.092990713649663, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.497101783752441, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8773550391197205, "num_tokens": 627714952.0, "step": 16453 }, { "epoch": 2.0931179239282534, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.471612930297852, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8641999959945679, "num_tokens": 627754710.0, "step": 16454 }, { "epoch": 2.093245134206844, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.464290618896484, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8783327341079712, "num_tokens": 627791504.0, "step": 16455 }, { "epoch": 2.0933723444854344, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.550368309020996, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8870483636856079, "num_tokens": 627829683.0, "step": 16456 }, { "epoch": 2.093499554764025, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.376304626464844, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8737919926643372, "num_tokens": 627869826.0, "step": 16457 }, { "epoch": 2.0936267650426155, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.510916709899902, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8707292079925537, "num_tokens": 627904165.0, "step": 16458 }, { "epoch": 2.093753975321206, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.481369972229004, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.878261923789978, "num_tokens": 627936559.0, "step": 16459 }, { "epoch": 2.0938811855997965, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.497549057006836, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8781910538673401, "num_tokens": 627971742.0, "step": 16460 }, { "epoch": 2.094008395878387, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.41019344329834, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8828341960906982, "num_tokens": 628013104.0, "step": 16461 }, { "epoch": 2.0941356061569776, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.496667861938477, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8701249957084656, "num_tokens": 628049378.0, "step": 16462 }, { "epoch": 2.094262816435568, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.46230411529541, "learning_rate": 1e-06, "loss": 0.3421, "mean_token_accuracy": 0.8917193412780762, "num_tokens": 628087213.0, "step": 16463 }, { "epoch": 2.0943900267141586, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.48176097869873, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8788901567459106, "num_tokens": 628122060.0, "step": 16464 }, { "epoch": 2.094517236992749, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.509559631347656, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8765838146209717, "num_tokens": 628166554.0, "step": 16465 }, { "epoch": 2.0946444472713397, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.526947021484375, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8714718222618103, "num_tokens": 628199953.0, "step": 16466 }, { "epoch": 2.09477165754993, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.450160026550293, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8770802021026611, "num_tokens": 628237044.0, "step": 16467 }, { "epoch": 2.0948988678285207, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.544877052307129, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8761504888534546, "num_tokens": 628275773.0, "step": 16468 }, { "epoch": 2.095026078107111, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.437124252319336, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.860131561756134, "num_tokens": 628314549.0, "step": 16469 }, { "epoch": 2.0951532883857014, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.49538516998291, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8700960874557495, "num_tokens": 628349417.0, "step": 16470 }, { "epoch": 2.095280498664292, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.487500190734863, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8777096271514893, "num_tokens": 628384435.0, "step": 16471 }, { "epoch": 2.0954077089428824, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.453878402709961, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8607979416847229, "num_tokens": 628424124.0, "step": 16472 }, { "epoch": 2.095534919221473, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.485498428344727, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8757100105285645, "num_tokens": 628462412.0, "step": 16473 }, { "epoch": 2.0956621295000635, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.506306648254395, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8559240698814392, "num_tokens": 628499066.0, "step": 16474 }, { "epoch": 2.095789339778654, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.504191398620605, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8873060941696167, "num_tokens": 628532379.0, "step": 16475 }, { "epoch": 2.0959165500572445, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.516451835632324, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8830209970474243, "num_tokens": 628570497.0, "step": 16476 }, { "epoch": 2.096043760335835, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.48678207397461, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8788599967956543, "num_tokens": 628608298.0, "step": 16477 }, { "epoch": 2.0961709706144256, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.510053634643555, "learning_rate": 1e-06, "loss": 0.3382, "mean_token_accuracy": 0.891343355178833, "num_tokens": 628641217.0, "step": 16478 }, { "epoch": 2.096298180893016, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.457905769348145, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8751957416534424, "num_tokens": 628682480.0, "step": 16479 }, { "epoch": 2.0964253911716066, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.449928283691406, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8721909523010254, "num_tokens": 628718359.0, "step": 16480 }, { "epoch": 2.096552601450197, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.512002944946289, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8802680373191833, "num_tokens": 628752214.0, "step": 16481 }, { "epoch": 2.0966798117287877, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.4951810836792, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8872370719909668, "num_tokens": 628787882.0, "step": 16482 }, { "epoch": 2.0968070220073782, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.47507095336914, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8708591461181641, "num_tokens": 628824910.0, "step": 16483 }, { "epoch": 2.0969342322859688, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.439857482910156, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8856331706047058, "num_tokens": 628860657.0, "step": 16484 }, { "epoch": 2.0970614425645593, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.497550010681152, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8578662872314453, "num_tokens": 628900619.0, "step": 16485 }, { "epoch": 2.09718865284315, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.460511207580566, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8861179351806641, "num_tokens": 628944304.0, "step": 16486 }, { "epoch": 2.0973158631217403, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.47098445892334, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8793814182281494, "num_tokens": 628980490.0, "step": 16487 }, { "epoch": 2.097443073400331, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.518726348876953, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8786863088607788, "num_tokens": 629011924.0, "step": 16488 }, { "epoch": 2.0975702836789214, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.457685470581055, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8494235277175903, "num_tokens": 629049514.0, "step": 16489 }, { "epoch": 2.097697493957512, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.511388778686523, "learning_rate": 1e-06, "loss": 0.3406, "mean_token_accuracy": 0.8920587301254272, "num_tokens": 629085374.0, "step": 16490 }, { "epoch": 2.0978247042361025, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.495018005371094, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8708779811859131, "num_tokens": 629126307.0, "step": 16491 }, { "epoch": 2.097951914514693, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.418188095092773, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8743051290512085, "num_tokens": 629159693.0, "step": 16492 }, { "epoch": 2.0980791247932835, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.509589195251465, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8801513910293579, "num_tokens": 629198544.0, "step": 16493 }, { "epoch": 2.0982063350718736, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.446159362792969, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8662484884262085, "num_tokens": 629240956.0, "step": 16494 }, { "epoch": 2.098333545350464, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.411385536193848, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8753359913825989, "num_tokens": 629276196.0, "step": 16495 }, { "epoch": 2.0984607556290547, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.497945785522461, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8615530729293823, "num_tokens": 629311398.0, "step": 16496 }, { "epoch": 2.098587965907645, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.391059875488281, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8687811493873596, "num_tokens": 629349488.0, "step": 16497 }, { "epoch": 2.0987151761862357, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.477730751037598, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8867530226707458, "num_tokens": 629388023.0, "step": 16498 }, { "epoch": 2.0988423864648262, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.447157859802246, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8728629350662231, "num_tokens": 629432140.0, "step": 16499 }, { "epoch": 2.0989695967434168, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.436476707458496, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8598660230636597, "num_tokens": 629480690.0, "step": 16500 }, { "epoch": 2.0990968070220073, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.396309852600098, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8913440704345703, "num_tokens": 629522500.0, "step": 16501 }, { "epoch": 2.099224017300598, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.465581893920898, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8743513226509094, "num_tokens": 629563450.0, "step": 16502 }, { "epoch": 2.0993512275791884, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.515145301818848, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8838717341423035, "num_tokens": 629603753.0, "step": 16503 }, { "epoch": 2.099478437857779, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.472259521484375, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8772721886634827, "num_tokens": 629637828.0, "step": 16504 }, { "epoch": 2.0996056481363694, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.452777862548828, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8814969658851624, "num_tokens": 629674636.0, "step": 16505 }, { "epoch": 2.09973285841496, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.459065437316895, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8647836446762085, "num_tokens": 629717046.0, "step": 16506 }, { "epoch": 2.0998600686935505, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.446026802062988, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8838814496994019, "num_tokens": 629750331.0, "step": 16507 }, { "epoch": 2.099987278972141, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.444175720214844, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8554915189743042, "num_tokens": 629790095.0, "step": 16508 }, { "epoch": 2.1001144892507315, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.493550300598145, "learning_rate": 1e-06, "loss": 0.3402, "mean_token_accuracy": 0.8880264759063721, "num_tokens": 629823035.0, "step": 16509 }, { "epoch": 2.100241699529322, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.503437995910645, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.870052695274353, "num_tokens": 629857096.0, "step": 16510 }, { "epoch": 2.1003689098079126, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.461382865905762, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8732141852378845, "num_tokens": 629893448.0, "step": 16511 }, { "epoch": 2.100496120086503, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.489289283752441, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8721061944961548, "num_tokens": 629931301.0, "step": 16512 }, { "epoch": 2.1006233303650936, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.48363971710205, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.882219672203064, "num_tokens": 629963481.0, "step": 16513 }, { "epoch": 2.100750540643684, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.490833282470703, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8700589537620544, "num_tokens": 630002739.0, "step": 16514 }, { "epoch": 2.1008777509222747, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.41884708404541, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8673176765441895, "num_tokens": 630041206.0, "step": 16515 }, { "epoch": 2.101004961200865, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.41889476776123, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8603182435035706, "num_tokens": 630083811.0, "step": 16516 }, { "epoch": 2.1011321714794557, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.423628807067871, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8714199662208557, "num_tokens": 630119823.0, "step": 16517 }, { "epoch": 2.1012593817580463, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.410526275634766, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8741588592529297, "num_tokens": 630163807.0, "step": 16518 }, { "epoch": 2.1013865920366364, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.478273391723633, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8688751459121704, "num_tokens": 630203038.0, "step": 16519 }, { "epoch": 2.101513802315227, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.465044021606445, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8783903121948242, "num_tokens": 630238505.0, "step": 16520 }, { "epoch": 2.1016410125938174, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.444375038146973, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8728201389312744, "num_tokens": 630275279.0, "step": 16521 }, { "epoch": 2.101768222872408, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.497071266174316, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8703347444534302, "num_tokens": 630317288.0, "step": 16522 }, { "epoch": 2.1018954331509985, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.417823791503906, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8741898536682129, "num_tokens": 630352786.0, "step": 16523 }, { "epoch": 2.102022643429589, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.516707420349121, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.874494731426239, "num_tokens": 630391138.0, "step": 16524 }, { "epoch": 2.1021498537081795, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.473506927490234, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8621730208396912, "num_tokens": 630426567.0, "step": 16525 }, { "epoch": 2.10227706398677, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.468269348144531, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8632688522338867, "num_tokens": 630467308.0, "step": 16526 }, { "epoch": 2.1024042742653606, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.504351615905762, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8846302032470703, "num_tokens": 630503333.0, "step": 16527 }, { "epoch": 2.102531484543951, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.52107048034668, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.86932373046875, "num_tokens": 630536838.0, "step": 16528 }, { "epoch": 2.1026586948225416, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.484227180480957, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8869446516036987, "num_tokens": 630570833.0, "step": 16529 }, { "epoch": 2.102785905101132, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.491362571716309, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.882465124130249, "num_tokens": 630615043.0, "step": 16530 }, { "epoch": 2.1029131153797227, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.486366271972656, "learning_rate": 1e-06, "loss": 0.3362, "mean_token_accuracy": 0.8960000276565552, "num_tokens": 630651368.0, "step": 16531 }, { "epoch": 2.1030403256583132, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.446263313293457, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.866951584815979, "num_tokens": 630690341.0, "step": 16532 }, { "epoch": 2.1031675359369038, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.55527400970459, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8713949918746948, "num_tokens": 630727168.0, "step": 16533 }, { "epoch": 2.1032947462154943, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.469719886779785, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8788238763809204, "num_tokens": 630766737.0, "step": 16534 }, { "epoch": 2.103421956494085, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.510293960571289, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8723632097244263, "num_tokens": 630803721.0, "step": 16535 }, { "epoch": 2.1035491667726753, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.477728843688965, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8805310130119324, "num_tokens": 630843697.0, "step": 16536 }, { "epoch": 2.103676377051266, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.479397773742676, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8758537173271179, "num_tokens": 630874325.0, "step": 16537 }, { "epoch": 2.1038035873298564, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.495750427246094, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8706583380699158, "num_tokens": 630911456.0, "step": 16538 }, { "epoch": 2.103930797608447, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.48198127746582, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8723082542419434, "num_tokens": 630954418.0, "step": 16539 }, { "epoch": 2.1040580078870375, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.385456085205078, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8734771609306335, "num_tokens": 630991826.0, "step": 16540 }, { "epoch": 2.104185218165628, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.504129409790039, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8740058541297913, "num_tokens": 631025831.0, "step": 16541 }, { "epoch": 2.1043124284442185, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.45870304107666, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8590640425682068, "num_tokens": 631062896.0, "step": 16542 }, { "epoch": 2.1044396387228086, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.457477569580078, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8801236748695374, "num_tokens": 631099291.0, "step": 16543 }, { "epoch": 2.104566849001399, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.51363754272461, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8661842346191406, "num_tokens": 631141570.0, "step": 16544 }, { "epoch": 2.1046940592799896, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.454797744750977, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8662638068199158, "num_tokens": 631178882.0, "step": 16545 }, { "epoch": 2.10482126955858, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.518790245056152, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8710160255432129, "num_tokens": 631218059.0, "step": 16546 }, { "epoch": 2.1049484798371707, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.500380516052246, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8757649064064026, "num_tokens": 631252827.0, "step": 16547 }, { "epoch": 2.1050756901157612, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.501120567321777, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8810032606124878, "num_tokens": 631294370.0, "step": 16548 }, { "epoch": 2.1052029003943518, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.456015586853027, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8700718283653259, "num_tokens": 631331091.0, "step": 16549 }, { "epoch": 2.1053301106729423, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.527639389038086, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8772687911987305, "num_tokens": 631367285.0, "step": 16550 }, { "epoch": 2.105457320951533, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.494356155395508, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8603399991989136, "num_tokens": 631402471.0, "step": 16551 }, { "epoch": 2.1055845312301233, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.556008338928223, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8738017082214355, "num_tokens": 631439706.0, "step": 16552 }, { "epoch": 2.105711741508714, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.519328117370605, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8784502744674683, "num_tokens": 631471437.0, "step": 16553 }, { "epoch": 2.1058389517873044, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.548502922058105, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8709492683410645, "num_tokens": 631515028.0, "step": 16554 }, { "epoch": 2.105966162065895, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.49829387664795, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8698973655700684, "num_tokens": 631552946.0, "step": 16555 }, { "epoch": 2.1060933723444855, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.516282081604004, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8787487149238586, "num_tokens": 631596045.0, "step": 16556 }, { "epoch": 2.106220582623076, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.511980056762695, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8818184733390808, "num_tokens": 631630233.0, "step": 16557 }, { "epoch": 2.1063477929016665, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.536505699157715, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8805102705955505, "num_tokens": 631666665.0, "step": 16558 }, { "epoch": 2.106475003180257, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.53965950012207, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.870802640914917, "num_tokens": 631709536.0, "step": 16559 }, { "epoch": 2.1066022134588476, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.52298641204834, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8883553743362427, "num_tokens": 631744333.0, "step": 16560 }, { "epoch": 2.106729423737438, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.562564849853516, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8602474927902222, "num_tokens": 631783522.0, "step": 16561 }, { "epoch": 2.1068566340160286, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.465147018432617, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8580694198608398, "num_tokens": 631822579.0, "step": 16562 }, { "epoch": 2.106983844294619, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.537994384765625, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8814809322357178, "num_tokens": 631862642.0, "step": 16563 }, { "epoch": 2.1071110545732097, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.539294242858887, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8845996260643005, "num_tokens": 631903161.0, "step": 16564 }, { "epoch": 2.1072382648518, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.540763854980469, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8774181604385376, "num_tokens": 631938149.0, "step": 16565 }, { "epoch": 2.1073654751303907, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.5823335647583, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8752954006195068, "num_tokens": 631974717.0, "step": 16566 }, { "epoch": 2.107492685408981, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.469106674194336, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.877936065196991, "num_tokens": 632011147.0, "step": 16567 }, { "epoch": 2.1076198956875714, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.512218475341797, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8797594308853149, "num_tokens": 632051915.0, "step": 16568 }, { "epoch": 2.107747105966162, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.539238929748535, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.864967942237854, "num_tokens": 632083047.0, "step": 16569 }, { "epoch": 2.1078743162447524, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.549595832824707, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8866196870803833, "num_tokens": 632119250.0, "step": 16570 }, { "epoch": 2.108001526523343, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.578818321228027, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8797551393508911, "num_tokens": 632152696.0, "step": 16571 }, { "epoch": 2.1081287368019335, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.4766845703125, "learning_rate": 1e-06, "loss": 0.337, "mean_token_accuracy": 0.8943641185760498, "num_tokens": 632188167.0, "step": 16572 }, { "epoch": 2.108255947080524, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.6141357421875, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.872609555721283, "num_tokens": 632229567.0, "step": 16573 }, { "epoch": 2.1083831573591145, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.517334938049316, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8836607336997986, "num_tokens": 632266870.0, "step": 16574 }, { "epoch": 2.108510367637705, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.600342750549316, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8739476799964905, "num_tokens": 632303844.0, "step": 16575 }, { "epoch": 2.1086375779162956, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.504666328430176, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.875511646270752, "num_tokens": 632339343.0, "step": 16576 }, { "epoch": 2.108764788194886, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.501812934875488, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8793308734893799, "num_tokens": 632385516.0, "step": 16577 }, { "epoch": 2.1088919984734766, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.530138969421387, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.878645122051239, "num_tokens": 632420664.0, "step": 16578 }, { "epoch": 2.109019208752067, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.525940895080566, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8796290159225464, "num_tokens": 632458082.0, "step": 16579 }, { "epoch": 2.1091464190306577, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.46582317352295, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8631206154823303, "num_tokens": 632495963.0, "step": 16580 }, { "epoch": 2.109273629309248, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.567179679870605, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8761383295059204, "num_tokens": 632529968.0, "step": 16581 }, { "epoch": 2.1094008395878387, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.554498672485352, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8824807405471802, "num_tokens": 632567527.0, "step": 16582 }, { "epoch": 2.1095280498664293, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.533666610717773, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8736415505409241, "num_tokens": 632606192.0, "step": 16583 }, { "epoch": 2.10965526014502, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.511293411254883, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8844594955444336, "num_tokens": 632645940.0, "step": 16584 }, { "epoch": 2.1097824704236103, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.513445854187012, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.882716178894043, "num_tokens": 632689385.0, "step": 16585 }, { "epoch": 2.109909680702201, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.54715633392334, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.881041944026947, "num_tokens": 632720178.0, "step": 16586 }, { "epoch": 2.1100368909807914, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.493880271911621, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8669917583465576, "num_tokens": 632763249.0, "step": 16587 }, { "epoch": 2.110164101259382, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.52453327178955, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8563621044158936, "num_tokens": 632803293.0, "step": 16588 }, { "epoch": 2.1102913115379724, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.557479858398438, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8660855889320374, "num_tokens": 632838731.0, "step": 16589 }, { "epoch": 2.110418521816563, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.517918586730957, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8657412528991699, "num_tokens": 632875044.0, "step": 16590 }, { "epoch": 2.1105457320951535, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.633566856384277, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8544381856918335, "num_tokens": 632911330.0, "step": 16591 }, { "epoch": 2.1106729423737436, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.444948196411133, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8634380102157593, "num_tokens": 632952122.0, "step": 16592 }, { "epoch": 2.110800152652334, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.658854484558105, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8568897843360901, "num_tokens": 632985602.0, "step": 16593 }, { "epoch": 2.1109273629309246, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.480522155761719, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8751955628395081, "num_tokens": 633023792.0, "step": 16594 }, { "epoch": 2.111054573209515, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.576799392700195, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8755322098731995, "num_tokens": 633065457.0, "step": 16595 }, { "epoch": 2.1111817834881057, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.563496589660645, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8837911486625671, "num_tokens": 633101402.0, "step": 16596 }, { "epoch": 2.1113089937666962, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.477872848510742, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8658199906349182, "num_tokens": 633147002.0, "step": 16597 }, { "epoch": 2.1114362040452868, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.477340698242188, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8737584948539734, "num_tokens": 633184209.0, "step": 16598 }, { "epoch": 2.1115634143238773, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.587624549865723, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8781548142433167, "num_tokens": 633224525.0, "step": 16599 }, { "epoch": 2.111690624602468, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.505620002746582, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8895905017852783, "num_tokens": 633253171.0, "step": 16600 }, { "epoch": 2.1118178348810583, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.494014739990234, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8716429471969604, "num_tokens": 633294326.0, "step": 16601 }, { "epoch": 2.111945045159649, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.511856079101562, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8876510262489319, "num_tokens": 633327236.0, "step": 16602 }, { "epoch": 2.1120722554382394, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.445226669311523, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8842391967773438, "num_tokens": 633368593.0, "step": 16603 }, { "epoch": 2.11219946571683, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.502324104309082, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8526303172111511, "num_tokens": 633402598.0, "step": 16604 }, { "epoch": 2.1123266759954205, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.48373794555664, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8690515756607056, "num_tokens": 633437069.0, "step": 16605 }, { "epoch": 2.112453886274011, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.470848083496094, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8772760629653931, "num_tokens": 633479752.0, "step": 16606 }, { "epoch": 2.1125810965526015, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.540202140808105, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8705246448516846, "num_tokens": 633516949.0, "step": 16607 }, { "epoch": 2.112708306831192, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.449067115783691, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8782870769500732, "num_tokens": 633547511.0, "step": 16608 }, { "epoch": 2.1128355171097826, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.501172065734863, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8761125206947327, "num_tokens": 633583940.0, "step": 16609 }, { "epoch": 2.112962727388373, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.429409980773926, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8638564348220825, "num_tokens": 633626245.0, "step": 16610 }, { "epoch": 2.1130899376669636, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.513528823852539, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8768245577812195, "num_tokens": 633662889.0, "step": 16611 }, { "epoch": 2.113217147945554, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.495071411132812, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.870094358921051, "num_tokens": 633701461.0, "step": 16612 }, { "epoch": 2.1133443582241447, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.493982315063477, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8758721947669983, "num_tokens": 633741302.0, "step": 16613 }, { "epoch": 2.113471568502735, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.476240158081055, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8790225386619568, "num_tokens": 633780337.0, "step": 16614 }, { "epoch": 2.1135987787813257, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.56269359588623, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8815081715583801, "num_tokens": 633814450.0, "step": 16615 }, { "epoch": 2.1137259890599163, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.459433555603027, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8667738437652588, "num_tokens": 633856252.0, "step": 16616 }, { "epoch": 2.1138531993385064, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.520488739013672, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8624988794326782, "num_tokens": 633895235.0, "step": 16617 }, { "epoch": 2.113980409617097, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.474411964416504, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8758400678634644, "num_tokens": 633938948.0, "step": 16618 }, { "epoch": 2.1141076198956874, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.523322105407715, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8808083534240723, "num_tokens": 633973393.0, "step": 16619 }, { "epoch": 2.114234830174278, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.487161636352539, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8782747983932495, "num_tokens": 634006714.0, "step": 16620 }, { "epoch": 2.1143620404528685, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.525006294250488, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8793890476226807, "num_tokens": 634046104.0, "step": 16621 }, { "epoch": 2.114489250731459, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.498710632324219, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8666254281997681, "num_tokens": 634086401.0, "step": 16622 }, { "epoch": 2.1146164610100495, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.5355224609375, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8662901520729065, "num_tokens": 634125479.0, "step": 16623 }, { "epoch": 2.11474367128864, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.530192375183105, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8561123609542847, "num_tokens": 634163873.0, "step": 16624 }, { "epoch": 2.1148708815672306, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.523518562316895, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8720378875732422, "num_tokens": 634200533.0, "step": 16625 }, { "epoch": 2.114998091845821, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.526899337768555, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8669955730438232, "num_tokens": 634235891.0, "step": 16626 }, { "epoch": 2.1151253021244116, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.491353034973145, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8732094168663025, "num_tokens": 634266748.0, "step": 16627 }, { "epoch": 2.115252512403002, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.57311725616455, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8628369569778442, "num_tokens": 634302603.0, "step": 16628 }, { "epoch": 2.1153797226815927, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.468995094299316, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8756248354911804, "num_tokens": 634345471.0, "step": 16629 }, { "epoch": 2.115506932960183, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.535760879516602, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8719937801361084, "num_tokens": 634383902.0, "step": 16630 }, { "epoch": 2.1156341432387737, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.56275749206543, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8677831888198853, "num_tokens": 634421082.0, "step": 16631 }, { "epoch": 2.1157613535173643, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.540108680725098, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8680477738380432, "num_tokens": 634456490.0, "step": 16632 }, { "epoch": 2.115888563795955, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.522801399230957, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8559437990188599, "num_tokens": 634498252.0, "step": 16633 }, { "epoch": 2.1160157740745453, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.535110473632812, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.881501317024231, "num_tokens": 634540769.0, "step": 16634 }, { "epoch": 2.116142984353136, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.570490837097168, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8810662627220154, "num_tokens": 634573807.0, "step": 16635 }, { "epoch": 2.1162701946317264, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.525317192077637, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8699995279312134, "num_tokens": 634609542.0, "step": 16636 }, { "epoch": 2.116397404910317, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.530780792236328, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8820885419845581, "num_tokens": 634648641.0, "step": 16637 }, { "epoch": 2.1165246151889074, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.48698616027832, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8725605010986328, "num_tokens": 634685531.0, "step": 16638 }, { "epoch": 2.116651825467498, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.646888732910156, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8749853372573853, "num_tokens": 634717239.0, "step": 16639 }, { "epoch": 2.116779035746088, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.553304672241211, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8710107207298279, "num_tokens": 634750058.0, "step": 16640 }, { "epoch": 2.1169062460246786, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.547576904296875, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8669790625572205, "num_tokens": 634793480.0, "step": 16641 }, { "epoch": 2.117033456303269, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.5609130859375, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8638962507247925, "num_tokens": 634829923.0, "step": 16642 }, { "epoch": 2.1171606665818596, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.505922317504883, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8685308694839478, "num_tokens": 634868499.0, "step": 16643 }, { "epoch": 2.11728787686045, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.580365180969238, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8822439908981323, "num_tokens": 634905922.0, "step": 16644 }, { "epoch": 2.1174150871390407, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.566929817199707, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8739800453186035, "num_tokens": 634946190.0, "step": 16645 }, { "epoch": 2.1175422974176312, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.593031883239746, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8855623602867126, "num_tokens": 634981840.0, "step": 16646 }, { "epoch": 2.1176695076962218, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.578569412231445, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8719280958175659, "num_tokens": 635024755.0, "step": 16647 }, { "epoch": 2.1177967179748123, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.55293083190918, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8777111768722534, "num_tokens": 635063076.0, "step": 16648 }, { "epoch": 2.117923928253403, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.570980072021484, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8858148455619812, "num_tokens": 635103052.0, "step": 16649 }, { "epoch": 2.1180511385319933, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.603365898132324, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8827165365219116, "num_tokens": 635140917.0, "step": 16650 }, { "epoch": 2.118178348810584, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.48592472076416, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8783283233642578, "num_tokens": 635182049.0, "step": 16651 }, { "epoch": 2.1183055590891744, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.578176498413086, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8692224025726318, "num_tokens": 635217613.0, "step": 16652 }, { "epoch": 2.118432769367765, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.510515213012695, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8826436996459961, "num_tokens": 635252716.0, "step": 16653 }, { "epoch": 2.1185599796463555, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.571321487426758, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.876758337020874, "num_tokens": 635288919.0, "step": 16654 }, { "epoch": 2.118687189924946, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.570948600769043, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8695329427719116, "num_tokens": 635329781.0, "step": 16655 }, { "epoch": 2.1188144002035365, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.53649616241455, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8835031986236572, "num_tokens": 635363844.0, "step": 16656 }, { "epoch": 2.118941610482127, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.525426864624023, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8706061244010925, "num_tokens": 635397568.0, "step": 16657 }, { "epoch": 2.1190688207607176, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.567975997924805, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8585740327835083, "num_tokens": 635432903.0, "step": 16658 }, { "epoch": 2.119196031039308, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.48800277709961, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.875954270362854, "num_tokens": 635473388.0, "step": 16659 }, { "epoch": 2.1193232413178986, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.509976387023926, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8769803047180176, "num_tokens": 635512966.0, "step": 16660 }, { "epoch": 2.119450451596489, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.591118812561035, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.861149787902832, "num_tokens": 635547814.0, "step": 16661 }, { "epoch": 2.1195776618750797, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.545323371887207, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8635908365249634, "num_tokens": 635583120.0, "step": 16662 }, { "epoch": 2.11970487215367, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.572643280029297, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.868766188621521, "num_tokens": 635621531.0, "step": 16663 }, { "epoch": 2.1198320824322607, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.574337005615234, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8641840219497681, "num_tokens": 635659641.0, "step": 16664 }, { "epoch": 2.119959292710851, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.531109809875488, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8692631721496582, "num_tokens": 635697953.0, "step": 16665 }, { "epoch": 2.1200865029894413, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.501968383789062, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8481313586235046, "num_tokens": 635742519.0, "step": 16666 }, { "epoch": 2.120213713268032, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.487005233764648, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8812020421028137, "num_tokens": 635777786.0, "step": 16667 }, { "epoch": 2.1203409235466224, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.574995040893555, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8761709928512573, "num_tokens": 635814490.0, "step": 16668 }, { "epoch": 2.120468133825213, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.501193046569824, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8607919812202454, "num_tokens": 635858351.0, "step": 16669 }, { "epoch": 2.1205953441038035, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.644274711608887, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8719442486763, "num_tokens": 635895585.0, "step": 16670 }, { "epoch": 2.120722554382394, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.469779014587402, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8803189992904663, "num_tokens": 635933591.0, "step": 16671 }, { "epoch": 2.1208497646609845, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.549811363220215, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8706372380256653, "num_tokens": 635972040.0, "step": 16672 }, { "epoch": 2.120976974939575, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.535853385925293, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.877133846282959, "num_tokens": 636013287.0, "step": 16673 }, { "epoch": 2.1211041852181656, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.47485637664795, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8682354688644409, "num_tokens": 636046137.0, "step": 16674 }, { "epoch": 2.121231395496756, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.444540977478027, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8629529476165771, "num_tokens": 636088870.0, "step": 16675 }, { "epoch": 2.1213586057753466, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.547916412353516, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.883442223072052, "num_tokens": 636130519.0, "step": 16676 }, { "epoch": 2.121485816053937, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.510340690612793, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8771869540214539, "num_tokens": 636165641.0, "step": 16677 }, { "epoch": 2.1216130263325277, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.50939655303955, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8604087233543396, "num_tokens": 636203811.0, "step": 16678 }, { "epoch": 2.121740236611118, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.500279426574707, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8856821656227112, "num_tokens": 636239503.0, "step": 16679 }, { "epoch": 2.1218674468897087, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.524002075195312, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.876907229423523, "num_tokens": 636278994.0, "step": 16680 }, { "epoch": 2.1219946571682993, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.46757698059082, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8696744441986084, "num_tokens": 636318025.0, "step": 16681 }, { "epoch": 2.12212186744689, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.514856338500977, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8750778436660767, "num_tokens": 636357289.0, "step": 16682 }, { "epoch": 2.1222490777254803, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.52912712097168, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8583414554595947, "num_tokens": 636399067.0, "step": 16683 }, { "epoch": 2.122376288004071, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.464461326599121, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8788461685180664, "num_tokens": 636442418.0, "step": 16684 }, { "epoch": 2.1225034982826614, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.549942970275879, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8622615933418274, "num_tokens": 636484273.0, "step": 16685 }, { "epoch": 2.122630708561252, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.514840126037598, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8829984664916992, "num_tokens": 636525675.0, "step": 16686 }, { "epoch": 2.1227579188398424, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.538617134094238, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.871578574180603, "num_tokens": 636559652.0, "step": 16687 }, { "epoch": 2.122885129118433, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.505194664001465, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.891474187374115, "num_tokens": 636597900.0, "step": 16688 }, { "epoch": 2.1230123393970235, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.536015510559082, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8752027750015259, "num_tokens": 636635822.0, "step": 16689 }, { "epoch": 2.1231395496756136, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.615702629089355, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.863921046257019, "num_tokens": 636667275.0, "step": 16690 }, { "epoch": 2.123266759954204, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.57376766204834, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8758597373962402, "num_tokens": 636704723.0, "step": 16691 }, { "epoch": 2.1233939702327946, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.507783889770508, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8784177303314209, "num_tokens": 636748209.0, "step": 16692 }, { "epoch": 2.123521180511385, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.541136741638184, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8787751197814941, "num_tokens": 636785052.0, "step": 16693 }, { "epoch": 2.1236483907899757, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.532333374023438, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8859530091285706, "num_tokens": 636816073.0, "step": 16694 }, { "epoch": 2.123775601068566, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.54202651977539, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8812823295593262, "num_tokens": 636848963.0, "step": 16695 }, { "epoch": 2.1239028113471567, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.45910358428955, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8864367008209229, "num_tokens": 636880632.0, "step": 16696 }, { "epoch": 2.1240300216257473, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.557684898376465, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.861624538898468, "num_tokens": 636923597.0, "step": 16697 }, { "epoch": 2.124157231904338, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.440393447875977, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8684507608413696, "num_tokens": 636962250.0, "step": 16698 }, { "epoch": 2.1242844421829283, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.481866836547852, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8807876110076904, "num_tokens": 636997973.0, "step": 16699 }, { "epoch": 2.124411652461519, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.599621772766113, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8618658781051636, "num_tokens": 637037320.0, "step": 16700 }, { "epoch": 2.1245388627401094, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.522454261779785, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8521262407302856, "num_tokens": 637072267.0, "step": 16701 }, { "epoch": 2.1246660730187, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.533053398132324, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8880137205123901, "num_tokens": 637108782.0, "step": 16702 }, { "epoch": 2.1247932832972904, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.564826011657715, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.873256266117096, "num_tokens": 637147368.0, "step": 16703 }, { "epoch": 2.124920493575881, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.541288375854492, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8824243545532227, "num_tokens": 637181909.0, "step": 16704 }, { "epoch": 2.1250477038544715, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.504024505615234, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8745651245117188, "num_tokens": 637218498.0, "step": 16705 }, { "epoch": 2.125174914133062, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.637243270874023, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8750183582305908, "num_tokens": 637254825.0, "step": 16706 }, { "epoch": 2.1253021244116526, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.470391273498535, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8643485903739929, "num_tokens": 637301593.0, "step": 16707 }, { "epoch": 2.125429334690243, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.577138900756836, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8854748010635376, "num_tokens": 637341676.0, "step": 16708 }, { "epoch": 2.1255565449688336, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.538294792175293, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8782955408096313, "num_tokens": 637378840.0, "step": 16709 }, { "epoch": 2.125683755247424, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.577061653137207, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.856423020362854, "num_tokens": 637412711.0, "step": 16710 }, { "epoch": 2.1258109655260147, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.584115982055664, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8726097345352173, "num_tokens": 637448287.0, "step": 16711 }, { "epoch": 2.125938175804605, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.502309799194336, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8716275691986084, "num_tokens": 637487899.0, "step": 16712 }, { "epoch": 2.1260653860831957, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.555256843566895, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8788917064666748, "num_tokens": 637528161.0, "step": 16713 }, { "epoch": 2.1261925963617863, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.482259750366211, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8625530004501343, "num_tokens": 637565692.0, "step": 16714 }, { "epoch": 2.1263198066403763, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.497784614562988, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8607688546180725, "num_tokens": 637600368.0, "step": 16715 }, { "epoch": 2.126447016918967, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.58210563659668, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8710125684738159, "num_tokens": 637635732.0, "step": 16716 }, { "epoch": 2.1265742271975574, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.460742950439453, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8688404560089111, "num_tokens": 637677127.0, "step": 16717 }, { "epoch": 2.126701437476148, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.562817573547363, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8737604022026062, "num_tokens": 637715278.0, "step": 16718 }, { "epoch": 2.1268286477547385, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.560596466064453, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8771249055862427, "num_tokens": 637742937.0, "step": 16719 }, { "epoch": 2.126955858033329, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.502974510192871, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8758955001831055, "num_tokens": 637776761.0, "step": 16720 }, { "epoch": 2.1270830683119195, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.51356315612793, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8754949569702148, "num_tokens": 637812197.0, "step": 16721 }, { "epoch": 2.12721027859051, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.47888469696045, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8784629106521606, "num_tokens": 637850442.0, "step": 16722 }, { "epoch": 2.1273374888691006, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.57105541229248, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8746001720428467, "num_tokens": 637889352.0, "step": 16723 }, { "epoch": 2.127464699147691, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.459417343139648, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8563927412033081, "num_tokens": 637927195.0, "step": 16724 }, { "epoch": 2.1275919094262816, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.597024917602539, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8628190755844116, "num_tokens": 637969431.0, "step": 16725 }, { "epoch": 2.127719119704872, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.54425048828125, "learning_rate": 1e-06, "loss": 0.346, "mean_token_accuracy": 0.8905936479568481, "num_tokens": 638013910.0, "step": 16726 }, { "epoch": 2.1278463299834627, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.592789649963379, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8633536100387573, "num_tokens": 638051848.0, "step": 16727 }, { "epoch": 2.127973540262053, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.503495216369629, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8738002777099609, "num_tokens": 638092192.0, "step": 16728 }, { "epoch": 2.1281007505406437, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.562163352966309, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8824630975723267, "num_tokens": 638134306.0, "step": 16729 }, { "epoch": 2.1282279608192343, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.54356575012207, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8599748611450195, "num_tokens": 638177988.0, "step": 16730 }, { "epoch": 2.128355171097825, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.553946495056152, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8839793801307678, "num_tokens": 638212237.0, "step": 16731 }, { "epoch": 2.1284823813764153, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.52095890045166, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8667724132537842, "num_tokens": 638252492.0, "step": 16732 }, { "epoch": 2.128609591655006, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.498876571655273, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8733503818511963, "num_tokens": 638294308.0, "step": 16733 }, { "epoch": 2.1287368019335964, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.518290519714355, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8774835467338562, "num_tokens": 638330178.0, "step": 16734 }, { "epoch": 2.128864012212187, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.562846183776855, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8534470200538635, "num_tokens": 638372731.0, "step": 16735 }, { "epoch": 2.1289912224907774, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.56611442565918, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8693991899490356, "num_tokens": 638407522.0, "step": 16736 }, { "epoch": 2.129118432769368, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.565051078796387, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8758901357650757, "num_tokens": 638443262.0, "step": 16737 }, { "epoch": 2.129245643047958, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.6568603515625, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8732877969741821, "num_tokens": 638482729.0, "step": 16738 }, { "epoch": 2.129372853326549, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.554635047912598, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8815776109695435, "num_tokens": 638520034.0, "step": 16739 }, { "epoch": 2.129500063605139, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.546295166015625, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8808120489120483, "num_tokens": 638558631.0, "step": 16740 }, { "epoch": 2.1296272738837296, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.558945655822754, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8852212429046631, "num_tokens": 638593565.0, "step": 16741 }, { "epoch": 2.12975448416232, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.534488677978516, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8781847953796387, "num_tokens": 638625952.0, "step": 16742 }, { "epoch": 2.1298816944409107, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.640703201293945, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8707058429718018, "num_tokens": 638662000.0, "step": 16743 }, { "epoch": 2.130008904719501, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.551100730895996, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.860065221786499, "num_tokens": 638706020.0, "step": 16744 }, { "epoch": 2.1301361149980917, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.600751876831055, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.880020260810852, "num_tokens": 638745354.0, "step": 16745 }, { "epoch": 2.1302633252766823, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.51389217376709, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8676294088363647, "num_tokens": 638785497.0, "step": 16746 }, { "epoch": 2.130390535555273, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.579286575317383, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.853432297706604, "num_tokens": 638821650.0, "step": 16747 }, { "epoch": 2.1305177458338633, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.614622116088867, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8605780601501465, "num_tokens": 638855589.0, "step": 16748 }, { "epoch": 2.130644956112454, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.531929016113281, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8728808760643005, "num_tokens": 638895766.0, "step": 16749 }, { "epoch": 2.1307721663910444, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.561885833740234, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8638507127761841, "num_tokens": 638932622.0, "step": 16750 }, { "epoch": 2.130899376669635, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.533437728881836, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8638707399368286, "num_tokens": 638971588.0, "step": 16751 }, { "epoch": 2.1310265869482254, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.55058479309082, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8793136477470398, "num_tokens": 639009921.0, "step": 16752 }, { "epoch": 2.131153797226816, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.56187629699707, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8728337287902832, "num_tokens": 639048160.0, "step": 16753 }, { "epoch": 2.1312810075054065, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.5921049118042, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.885632336139679, "num_tokens": 639080260.0, "step": 16754 }, { "epoch": 2.131408217783997, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.594182968139648, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8663709163665771, "num_tokens": 639118758.0, "step": 16755 }, { "epoch": 2.1315354280625876, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.575209617614746, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8740030527114868, "num_tokens": 639159528.0, "step": 16756 }, { "epoch": 2.131662638341178, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.582842826843262, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8767551779747009, "num_tokens": 639200394.0, "step": 16757 }, { "epoch": 2.1317898486197686, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61500358581543, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8649757504463196, "num_tokens": 639240938.0, "step": 16758 }, { "epoch": 2.131917058898359, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.579777717590332, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.868160605430603, "num_tokens": 639278097.0, "step": 16759 }, { "epoch": 2.1320442691769497, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.49577522277832, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.872498631477356, "num_tokens": 639324512.0, "step": 16760 }, { "epoch": 2.13217147945554, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.596311569213867, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.878135085105896, "num_tokens": 639366833.0, "step": 16761 }, { "epoch": 2.1322986897341307, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.519206047058105, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8583950996398926, "num_tokens": 639405765.0, "step": 16762 }, { "epoch": 2.132425900012721, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.590750694274902, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8703594207763672, "num_tokens": 639439622.0, "step": 16763 }, { "epoch": 2.1325531102913113, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.584647178649902, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8839791417121887, "num_tokens": 639479943.0, "step": 16764 }, { "epoch": 2.132680320569902, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.5068941116333, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8685383796691895, "num_tokens": 639520487.0, "step": 16765 }, { "epoch": 2.1328075308484924, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.539562225341797, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8719874024391174, "num_tokens": 639558997.0, "step": 16766 }, { "epoch": 2.132934741127083, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.528985023498535, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.871692419052124, "num_tokens": 639594569.0, "step": 16767 }, { "epoch": 2.1330619514056735, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.536874771118164, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8868687748908997, "num_tokens": 639629827.0, "step": 16768 }, { "epoch": 2.133189161684264, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.587684631347656, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8677399158477783, "num_tokens": 639667634.0, "step": 16769 }, { "epoch": 2.1333163719628545, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.614018440246582, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.872143030166626, "num_tokens": 639700536.0, "step": 16770 }, { "epoch": 2.133443582241445, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.556439399719238, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.875218391418457, "num_tokens": 639744006.0, "step": 16771 }, { "epoch": 2.1335707925200356, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.547476768493652, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8788524866104126, "num_tokens": 639784950.0, "step": 16772 }, { "epoch": 2.133698002798626, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.536673545837402, "learning_rate": 1e-06, "loss": 0.3537, "mean_token_accuracy": 0.8867460489273071, "num_tokens": 639823567.0, "step": 16773 }, { "epoch": 2.1338252130772166, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.601520538330078, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8844391703605652, "num_tokens": 639863386.0, "step": 16774 }, { "epoch": 2.133952423355807, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.5341796875, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8619413375854492, "num_tokens": 639901219.0, "step": 16775 }, { "epoch": 2.1340796336343977, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.601994514465332, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8578281402587891, "num_tokens": 639935240.0, "step": 16776 }, { "epoch": 2.134206843912988, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.599685668945312, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8652633428573608, "num_tokens": 639975370.0, "step": 16777 }, { "epoch": 2.1343340541915787, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.549715042114258, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.854736328125, "num_tokens": 640020463.0, "step": 16778 }, { "epoch": 2.1344612644701693, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.615336418151855, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8680936694145203, "num_tokens": 640057142.0, "step": 16779 }, { "epoch": 2.13458847474876, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.499504089355469, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8703384399414062, "num_tokens": 640098879.0, "step": 16780 }, { "epoch": 2.1347156850273503, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.614742279052734, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8756281137466431, "num_tokens": 640136077.0, "step": 16781 }, { "epoch": 2.134842895305941, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.571024894714355, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8589988350868225, "num_tokens": 640175160.0, "step": 16782 }, { "epoch": 2.1349701055845314, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.476690292358398, "learning_rate": 1e-06, "loss": 0.329, "mean_token_accuracy": 0.897297739982605, "num_tokens": 640216670.0, "step": 16783 }, { "epoch": 2.135097315863122, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.622547149658203, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8702775239944458, "num_tokens": 640254872.0, "step": 16784 }, { "epoch": 2.1352245261417124, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.581265449523926, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8713273406028748, "num_tokens": 640289319.0, "step": 16785 }, { "epoch": 2.135351736420303, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.567076683044434, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8878941535949707, "num_tokens": 640323897.0, "step": 16786 }, { "epoch": 2.1354789466988935, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.510634422302246, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8788455724716187, "num_tokens": 640366203.0, "step": 16787 }, { "epoch": 2.1356061569774836, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.50377082824707, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8832594752311707, "num_tokens": 640408089.0, "step": 16788 }, { "epoch": 2.135733367256074, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.531423568725586, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8827428221702576, "num_tokens": 640443532.0, "step": 16789 }, { "epoch": 2.1358605775346646, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.575834274291992, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8667356371879578, "num_tokens": 640480425.0, "step": 16790 }, { "epoch": 2.135987787813255, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.470440864562988, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8753455281257629, "num_tokens": 640525658.0, "step": 16791 }, { "epoch": 2.1361149980918457, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.535979270935059, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8819658160209656, "num_tokens": 640563276.0, "step": 16792 }, { "epoch": 2.136242208370436, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.559345245361328, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.865855872631073, "num_tokens": 640608150.0, "step": 16793 }, { "epoch": 2.1363694186490267, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.593011856079102, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8758663535118103, "num_tokens": 640646127.0, "step": 16794 }, { "epoch": 2.1364966289276173, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.538946151733398, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8791409730911255, "num_tokens": 640683195.0, "step": 16795 }, { "epoch": 2.136623839206208, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.516304969787598, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8863540291786194, "num_tokens": 640724102.0, "step": 16796 }, { "epoch": 2.1367510494847983, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.539569854736328, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8701130151748657, "num_tokens": 640762624.0, "step": 16797 }, { "epoch": 2.136878259763389, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.603865623474121, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.868515133857727, "num_tokens": 640799437.0, "step": 16798 }, { "epoch": 2.1370054700419794, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.519447326660156, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8792303204536438, "num_tokens": 640837115.0, "step": 16799 }, { "epoch": 2.13713268032057, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.4940767288208, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.873429536819458, "num_tokens": 640882312.0, "step": 16800 }, { "epoch": 2.1372598905991604, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.600150108337402, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8731474280357361, "num_tokens": 640921032.0, "step": 16801 }, { "epoch": 2.137387100877751, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.530972480773926, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8742368221282959, "num_tokens": 640957424.0, "step": 16802 }, { "epoch": 2.1375143111563415, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.578152656555176, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8648300170898438, "num_tokens": 640992985.0, "step": 16803 }, { "epoch": 2.137641521434932, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.543314933776855, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8813648223876953, "num_tokens": 641030379.0, "step": 16804 }, { "epoch": 2.1377687317135226, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.533730506896973, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8653252124786377, "num_tokens": 641068526.0, "step": 16805 }, { "epoch": 2.137895941992113, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.535439491271973, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8671278953552246, "num_tokens": 641103351.0, "step": 16806 }, { "epoch": 2.1380231522707036, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.56525707244873, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8759728670120239, "num_tokens": 641143570.0, "step": 16807 }, { "epoch": 2.138150362549294, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.500373840332031, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8785565495491028, "num_tokens": 641180625.0, "step": 16808 }, { "epoch": 2.1382775728278847, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.586853981018066, "learning_rate": 1e-06, "loss": 0.4721, "mean_token_accuracy": 0.852791965007782, "num_tokens": 641226770.0, "step": 16809 }, { "epoch": 2.138404783106475, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.539900779724121, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8841986656188965, "num_tokens": 641269682.0, "step": 16810 }, { "epoch": 2.1385319933850653, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.580329895019531, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8862137794494629, "num_tokens": 641303803.0, "step": 16811 }, { "epoch": 2.1386592036636562, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.59164047241211, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8679503202438354, "num_tokens": 641343141.0, "step": 16812 }, { "epoch": 2.1387864139422463, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.540077209472656, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8776662945747375, "num_tokens": 641384006.0, "step": 16813 }, { "epoch": 2.138913624220837, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.565065383911133, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8529893755912781, "num_tokens": 641423998.0, "step": 16814 }, { "epoch": 2.1390408344994274, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.51553726196289, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8722265362739563, "num_tokens": 641457886.0, "step": 16815 }, { "epoch": 2.139168044778018, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.50243091583252, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8740005493164062, "num_tokens": 641503386.0, "step": 16816 }, { "epoch": 2.1392952550566084, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.58953857421875, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8836725950241089, "num_tokens": 641539533.0, "step": 16817 }, { "epoch": 2.139422465335199, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.584346771240234, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8687469363212585, "num_tokens": 641574020.0, "step": 16818 }, { "epoch": 2.1395496756137895, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.599127769470215, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8759000301361084, "num_tokens": 641605206.0, "step": 16819 }, { "epoch": 2.13967688589238, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.508612632751465, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8635727167129517, "num_tokens": 641649857.0, "step": 16820 }, { "epoch": 2.1398040961709706, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.605088233947754, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8689225912094116, "num_tokens": 641683360.0, "step": 16821 }, { "epoch": 2.139931306449561, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.569889068603516, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8736841678619385, "num_tokens": 641720105.0, "step": 16822 }, { "epoch": 2.1400585167281516, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.644818305969238, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8755452036857605, "num_tokens": 641750718.0, "step": 16823 }, { "epoch": 2.140185727006742, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.588157653808594, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8720290660858154, "num_tokens": 641791620.0, "step": 16824 }, { "epoch": 2.1403129372853327, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.563309669494629, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8620461225509644, "num_tokens": 641833566.0, "step": 16825 }, { "epoch": 2.140440147563923, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.678414344787598, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8792051076889038, "num_tokens": 641869080.0, "step": 16826 }, { "epoch": 2.1405673578425137, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.535093307495117, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8710908889770508, "num_tokens": 641910195.0, "step": 16827 }, { "epoch": 2.1406945681211043, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.599533081054688, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8601988554000854, "num_tokens": 641947270.0, "step": 16828 }, { "epoch": 2.140821778399695, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.619315147399902, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8747830390930176, "num_tokens": 641986472.0, "step": 16829 }, { "epoch": 2.1409489886782853, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.603384971618652, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8592041730880737, "num_tokens": 642028130.0, "step": 16830 }, { "epoch": 2.141076198956876, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.546024322509766, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8672968149185181, "num_tokens": 642067097.0, "step": 16831 }, { "epoch": 2.1412034092354664, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.701068878173828, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8847043514251709, "num_tokens": 642099628.0, "step": 16832 }, { "epoch": 2.141330619514057, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.573393821716309, "learning_rate": 1e-06, "loss": 0.3284, "mean_token_accuracy": 0.8954579830169678, "num_tokens": 642135941.0, "step": 16833 }, { "epoch": 2.1414578297926474, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.662307739257812, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8669841289520264, "num_tokens": 642169465.0, "step": 16834 }, { "epoch": 2.141585040071238, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.600618362426758, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8767145872116089, "num_tokens": 642208961.0, "step": 16835 }, { "epoch": 2.141712250349828, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.591894149780273, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.872211754322052, "num_tokens": 642249691.0, "step": 16836 }, { "epoch": 2.141839460628419, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.559743881225586, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8758621215820312, "num_tokens": 642292425.0, "step": 16837 }, { "epoch": 2.141966670907009, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.649413108825684, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8823100328445435, "num_tokens": 642330982.0, "step": 16838 }, { "epoch": 2.1420938811855996, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.532881736755371, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8749882578849792, "num_tokens": 642365414.0, "step": 16839 }, { "epoch": 2.14222109146419, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.595576286315918, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8676104545593262, "num_tokens": 642398660.0, "step": 16840 }, { "epoch": 2.1423483017427807, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.575226783752441, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8812386393547058, "num_tokens": 642432605.0, "step": 16841 }, { "epoch": 2.142475512021371, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.53974437713623, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8816213011741638, "num_tokens": 642472041.0, "step": 16842 }, { "epoch": 2.1426027222999617, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.600688934326172, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8835025429725647, "num_tokens": 642512399.0, "step": 16843 }, { "epoch": 2.1427299325785523, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.561565399169922, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8675091862678528, "num_tokens": 642550829.0, "step": 16844 }, { "epoch": 2.142857142857143, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.541494369506836, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8733525276184082, "num_tokens": 642591255.0, "step": 16845 }, { "epoch": 2.1429843531357333, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.62540054321289, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8756828308105469, "num_tokens": 642625772.0, "step": 16846 }, { "epoch": 2.143111563414324, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.550309181213379, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8669110536575317, "num_tokens": 642667596.0, "step": 16847 }, { "epoch": 2.1432387736929144, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.635298728942871, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8745119571685791, "num_tokens": 642702159.0, "step": 16848 }, { "epoch": 2.143365983971505, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.65489387512207, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8879320621490479, "num_tokens": 642737307.0, "step": 16849 }, { "epoch": 2.1434931942500954, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.610435485839844, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8636091947555542, "num_tokens": 642776660.0, "step": 16850 }, { "epoch": 2.143620404528686, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.613097190856934, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8671383857727051, "num_tokens": 642814131.0, "step": 16851 }, { "epoch": 2.1437476148072765, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.612065315246582, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8542794585227966, "num_tokens": 642854138.0, "step": 16852 }, { "epoch": 2.143874825085867, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.608291625976562, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8805463910102844, "num_tokens": 642893767.0, "step": 16853 }, { "epoch": 2.1440020353644575, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.597423553466797, "learning_rate": 1e-06, "loss": 0.347, "mean_token_accuracy": 0.8868238925933838, "num_tokens": 642928286.0, "step": 16854 }, { "epoch": 2.144129245643048, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.593791961669922, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8809245824813843, "num_tokens": 642971220.0, "step": 16855 }, { "epoch": 2.1442564559216386, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.631550788879395, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8774571418762207, "num_tokens": 643013618.0, "step": 16856 }, { "epoch": 2.144383666200229, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.5393648147583, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.876960277557373, "num_tokens": 643058122.0, "step": 16857 }, { "epoch": 2.1445108764788197, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.620720863342285, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8614870309829712, "num_tokens": 643096662.0, "step": 16858 }, { "epoch": 2.14463808675741, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.537907600402832, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8853392601013184, "num_tokens": 643137923.0, "step": 16859 }, { "epoch": 2.1447652970360007, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.556905746459961, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8789347410202026, "num_tokens": 643176918.0, "step": 16860 }, { "epoch": 2.144892507314591, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.65230941772461, "learning_rate": 1e-06, "loss": 0.3466, "mean_token_accuracy": 0.8922243118286133, "num_tokens": 643217525.0, "step": 16861 }, { "epoch": 2.1450197175931813, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.632366180419922, "learning_rate": 1e-06, "loss": 0.3436, "mean_token_accuracy": 0.8900806307792664, "num_tokens": 643255131.0, "step": 16862 }, { "epoch": 2.145146927871772, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.623977661132812, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8815947771072388, "num_tokens": 643292092.0, "step": 16863 }, { "epoch": 2.1452741381503624, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.591612815856934, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.870869517326355, "num_tokens": 643330510.0, "step": 16864 }, { "epoch": 2.145401348428953, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.580625534057617, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8691540956497192, "num_tokens": 643372900.0, "step": 16865 }, { "epoch": 2.1455285587075434, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.57072639465332, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8815803527832031, "num_tokens": 643412866.0, "step": 16866 }, { "epoch": 2.145655768986134, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.627145767211914, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8581241369247437, "num_tokens": 643452846.0, "step": 16867 }, { "epoch": 2.1457829792647245, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.656335830688477, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.879552960395813, "num_tokens": 643488264.0, "step": 16868 }, { "epoch": 2.145910189543315, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.561138153076172, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.862331748008728, "num_tokens": 643528131.0, "step": 16869 }, { "epoch": 2.1460373998219056, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.542315483093262, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.873851478099823, "num_tokens": 643564041.0, "step": 16870 }, { "epoch": 2.146164610100496, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.667027473449707, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.861956775188446, "num_tokens": 643601507.0, "step": 16871 }, { "epoch": 2.1462918203790866, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.457298278808594, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8582476377487183, "num_tokens": 643647161.0, "step": 16872 }, { "epoch": 2.146419030657677, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.582894325256348, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8793742656707764, "num_tokens": 643687114.0, "step": 16873 }, { "epoch": 2.1465462409362677, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.554973602294922, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8598586320877075, "num_tokens": 643722110.0, "step": 16874 }, { "epoch": 2.146673451214858, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.591581344604492, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8520926833152771, "num_tokens": 643765770.0, "step": 16875 }, { "epoch": 2.1468006614934487, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.555184364318848, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8659496307373047, "num_tokens": 643810097.0, "step": 16876 }, { "epoch": 2.1469278717720393, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.542278289794922, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8480315208435059, "num_tokens": 643847904.0, "step": 16877 }, { "epoch": 2.14705508205063, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.585428237915039, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8821206092834473, "num_tokens": 643889157.0, "step": 16878 }, { "epoch": 2.1471822923292203, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61473560333252, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8759801387786865, "num_tokens": 643931181.0, "step": 16879 }, { "epoch": 2.147309502607811, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.636906623840332, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8721428513526917, "num_tokens": 643965415.0, "step": 16880 }, { "epoch": 2.1474367128864014, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.517730712890625, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.85917067527771, "num_tokens": 644010647.0, "step": 16881 }, { "epoch": 2.147563923164992, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.616558074951172, "learning_rate": 1e-06, "loss": 0.3491, "mean_token_accuracy": 0.8894700407981873, "num_tokens": 644044347.0, "step": 16882 }, { "epoch": 2.1476911334435824, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.532563209533691, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8810850977897644, "num_tokens": 644083456.0, "step": 16883 }, { "epoch": 2.147818343722173, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.615565299987793, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8634470701217651, "num_tokens": 644123851.0, "step": 16884 }, { "epoch": 2.1479455540007635, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.494853019714355, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8827663660049438, "num_tokens": 644168880.0, "step": 16885 }, { "epoch": 2.1480727642793536, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.558467864990234, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8758547306060791, "num_tokens": 644204532.0, "step": 16886 }, { "epoch": 2.148199974557944, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.55260181427002, "learning_rate": 1e-06, "loss": 0.3415, "mean_token_accuracy": 0.8918970227241516, "num_tokens": 644246160.0, "step": 16887 }, { "epoch": 2.1483271848365346, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.531375885009766, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8857516646385193, "num_tokens": 644284881.0, "step": 16888 }, { "epoch": 2.148454395115125, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.528390884399414, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8670265674591064, "num_tokens": 644326661.0, "step": 16889 }, { "epoch": 2.1485816053937157, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.509976387023926, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8789764642715454, "num_tokens": 644364420.0, "step": 16890 }, { "epoch": 2.148708815672306, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.572988510131836, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8569352626800537, "num_tokens": 644398626.0, "step": 16891 }, { "epoch": 2.1488360259508967, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.563305854797363, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8743393421173096, "num_tokens": 644434772.0, "step": 16892 }, { "epoch": 2.1489632362294873, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.54224967956543, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8713980317115784, "num_tokens": 644474425.0, "step": 16893 }, { "epoch": 2.149090446508078, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.64340591430664, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8664686679840088, "num_tokens": 644507796.0, "step": 16894 }, { "epoch": 2.1492176567866683, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.56745719909668, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8698103427886963, "num_tokens": 644544172.0, "step": 16895 }, { "epoch": 2.149344867065259, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.595925331115723, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8729225397109985, "num_tokens": 644582217.0, "step": 16896 }, { "epoch": 2.1494720773438494, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.613794326782227, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8797898292541504, "num_tokens": 644620028.0, "step": 16897 }, { "epoch": 2.14959928762244, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.520563125610352, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8578263521194458, "num_tokens": 644663743.0, "step": 16898 }, { "epoch": 2.1497264979010304, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61025333404541, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8752771615982056, "num_tokens": 644700914.0, "step": 16899 }, { "epoch": 2.149853708179621, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.572720527648926, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8703993558883667, "num_tokens": 644744232.0, "step": 16900 }, { "epoch": 2.1499809184582115, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.62397289276123, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8671755790710449, "num_tokens": 644780851.0, "step": 16901 }, { "epoch": 2.150108128736802, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.559341430664062, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8820187449455261, "num_tokens": 644822103.0, "step": 16902 }, { "epoch": 2.1502353390153925, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.598388671875, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8712146282196045, "num_tokens": 644861209.0, "step": 16903 }, { "epoch": 2.150362549293983, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.601214408874512, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8610223531723022, "num_tokens": 644899860.0, "step": 16904 }, { "epoch": 2.1504897595725736, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.563152313232422, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8688650131225586, "num_tokens": 644943101.0, "step": 16905 }, { "epoch": 2.150616969851164, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.569786071777344, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8626354932785034, "num_tokens": 644981786.0, "step": 16906 }, { "epoch": 2.1507441801297547, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.567341804504395, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8750700950622559, "num_tokens": 645017949.0, "step": 16907 }, { "epoch": 2.150871390408345, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.577930450439453, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8720798492431641, "num_tokens": 645053443.0, "step": 16908 }, { "epoch": 2.1509986006869353, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.607134819030762, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.875264585018158, "num_tokens": 645089916.0, "step": 16909 }, { "epoch": 2.1511258109655262, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.584896087646484, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8569118976593018, "num_tokens": 645134531.0, "step": 16910 }, { "epoch": 2.1512530212441163, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.587119102478027, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.86842942237854, "num_tokens": 645172608.0, "step": 16911 }, { "epoch": 2.151380231522707, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.581976890563965, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8690645694732666, "num_tokens": 645209895.0, "step": 16912 }, { "epoch": 2.1515074418012974, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.555418968200684, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8787206411361694, "num_tokens": 645245259.0, "step": 16913 }, { "epoch": 2.151634652079888, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.555635452270508, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8681231737136841, "num_tokens": 645284962.0, "step": 16914 }, { "epoch": 2.1517618623584784, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.583695411682129, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8878955841064453, "num_tokens": 645322352.0, "step": 16915 }, { "epoch": 2.151889072637069, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.674637794494629, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8717032670974731, "num_tokens": 645358401.0, "step": 16916 }, { "epoch": 2.1520162829156595, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.583060264587402, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8849648833274841, "num_tokens": 645399147.0, "step": 16917 }, { "epoch": 2.15214349319425, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.581657409667969, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8548579216003418, "num_tokens": 645437286.0, "step": 16918 }, { "epoch": 2.1522707034728406, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.570597648620605, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.865881085395813, "num_tokens": 645477853.0, "step": 16919 }, { "epoch": 2.152397913751431, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.588888168334961, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8632413148880005, "num_tokens": 645520476.0, "step": 16920 }, { "epoch": 2.1525251240300216, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.68940544128418, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8589776754379272, "num_tokens": 645561539.0, "step": 16921 }, { "epoch": 2.152652334308612, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.524646759033203, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8674107789993286, "num_tokens": 645602950.0, "step": 16922 }, { "epoch": 2.1527795445872027, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.565114974975586, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8621441125869751, "num_tokens": 645639215.0, "step": 16923 }, { "epoch": 2.152906754865793, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.539624214172363, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8716337084770203, "num_tokens": 645677287.0, "step": 16924 }, { "epoch": 2.1530339651443837, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.541921615600586, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8800593614578247, "num_tokens": 645710784.0, "step": 16925 }, { "epoch": 2.1531611754229742, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.588814735412598, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8687889575958252, "num_tokens": 645750568.0, "step": 16926 }, { "epoch": 2.1532883857015648, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.595010757446289, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8778151273727417, "num_tokens": 645783911.0, "step": 16927 }, { "epoch": 2.1534155959801553, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61274528503418, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8890535831451416, "num_tokens": 645822524.0, "step": 16928 }, { "epoch": 2.153542806258746, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.577978134155273, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8695715665817261, "num_tokens": 645856924.0, "step": 16929 }, { "epoch": 2.1536700165373364, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.615669250488281, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.86537766456604, "num_tokens": 645896995.0, "step": 16930 }, { "epoch": 2.153797226815927, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.574252128601074, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8796344995498657, "num_tokens": 645941768.0, "step": 16931 }, { "epoch": 2.1539244370945174, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.544743537902832, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8711998462677002, "num_tokens": 645981909.0, "step": 16932 }, { "epoch": 2.154051647373108, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.593167304992676, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8642449378967285, "num_tokens": 646022034.0, "step": 16933 }, { "epoch": 2.154178857651698, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.568743705749512, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8716263771057129, "num_tokens": 646056714.0, "step": 16934 }, { "epoch": 2.1543060679302886, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.590901374816895, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8650017976760864, "num_tokens": 646095668.0, "step": 16935 }, { "epoch": 2.154433278208879, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.525002479553223, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.856269896030426, "num_tokens": 646136745.0, "step": 16936 }, { "epoch": 2.1545604884874696, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.593923568725586, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8692579865455627, "num_tokens": 646180314.0, "step": 16937 }, { "epoch": 2.15468769876606, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.570297241210938, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8743756413459778, "num_tokens": 646218191.0, "step": 16938 }, { "epoch": 2.1548149090446507, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.59920597076416, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8879170417785645, "num_tokens": 646256350.0, "step": 16939 }, { "epoch": 2.154942119323241, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.551067352294922, "learning_rate": 1e-06, "loss": 0.3215, "mean_token_accuracy": 0.8944666981697083, "num_tokens": 646290915.0, "step": 16940 }, { "epoch": 2.1550693296018317, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.599539756774902, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.865681529045105, "num_tokens": 646325838.0, "step": 16941 }, { "epoch": 2.1551965398804223, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.71908187866211, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8550795316696167, "num_tokens": 646362742.0, "step": 16942 }, { "epoch": 2.155323750159013, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.631880760192871, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8591791391372681, "num_tokens": 646403958.0, "step": 16943 }, { "epoch": 2.1554509604376033, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.63544750213623, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8665192723274231, "num_tokens": 646442872.0, "step": 16944 }, { "epoch": 2.155578170716194, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.559879302978516, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8794077634811401, "num_tokens": 646484642.0, "step": 16945 }, { "epoch": 2.1557053809947844, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.647543907165527, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8818279504776001, "num_tokens": 646521688.0, "step": 16946 }, { "epoch": 2.155832591273375, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.658220291137695, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8796883821487427, "num_tokens": 646560182.0, "step": 16947 }, { "epoch": 2.1559598015519654, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.605681419372559, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8771369457244873, "num_tokens": 646593252.0, "step": 16948 }, { "epoch": 2.156087011830556, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.677225112915039, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8703317642211914, "num_tokens": 646635514.0, "step": 16949 }, { "epoch": 2.1562142221091465, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.6911039352417, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8599257469177246, "num_tokens": 646674785.0, "step": 16950 }, { "epoch": 2.156341432387737, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.680562019348145, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8720245361328125, "num_tokens": 646713832.0, "step": 16951 }, { "epoch": 2.1564686426663275, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.576445579528809, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8761101961135864, "num_tokens": 646752218.0, "step": 16952 }, { "epoch": 2.156595852944918, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.748827934265137, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8711225390434265, "num_tokens": 646786196.0, "step": 16953 }, { "epoch": 2.1567230632235086, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.642719268798828, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8659576177597046, "num_tokens": 646821462.0, "step": 16954 }, { "epoch": 2.156850273502099, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.623274803161621, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8798172473907471, "num_tokens": 646856613.0, "step": 16955 }, { "epoch": 2.1569774837806897, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.592048645019531, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8845566511154175, "num_tokens": 646893013.0, "step": 16956 }, { "epoch": 2.15710469405928, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.656786918640137, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8772456645965576, "num_tokens": 646936180.0, "step": 16957 }, { "epoch": 2.1572319043378707, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.622498512268066, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8720505833625793, "num_tokens": 646979314.0, "step": 16958 }, { "epoch": 2.157359114616461, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.618444442749023, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8824818134307861, "num_tokens": 647020136.0, "step": 16959 }, { "epoch": 2.1574863248950513, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.66679859161377, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8768428564071655, "num_tokens": 647052970.0, "step": 16960 }, { "epoch": 2.157613535173642, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.647175788879395, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8665897846221924, "num_tokens": 647093195.0, "step": 16961 }, { "epoch": 2.1577407454522324, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.643044471740723, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8724455237388611, "num_tokens": 647133800.0, "step": 16962 }, { "epoch": 2.157867955730823, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.5787992477417, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8571231365203857, "num_tokens": 647167704.0, "step": 16963 }, { "epoch": 2.1579951660094134, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.657758712768555, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8599167466163635, "num_tokens": 647208474.0, "step": 16964 }, { "epoch": 2.158122376288004, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.696253776550293, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8850938677787781, "num_tokens": 647242131.0, "step": 16965 }, { "epoch": 2.1582495865665945, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.642837524414062, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8545501232147217, "num_tokens": 647280952.0, "step": 16966 }, { "epoch": 2.158376796845185, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.62440299987793, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8674371242523193, "num_tokens": 647318411.0, "step": 16967 }, { "epoch": 2.1585040071237755, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.613601684570312, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8598737716674805, "num_tokens": 647361002.0, "step": 16968 }, { "epoch": 2.158631217402366, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.652527809143066, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8757923245429993, "num_tokens": 647402995.0, "step": 16969 }, { "epoch": 2.1587584276809566, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.670742988586426, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8743636012077332, "num_tokens": 647441220.0, "step": 16970 }, { "epoch": 2.158885637959547, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.629741668701172, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.865566611289978, "num_tokens": 647476832.0, "step": 16971 }, { "epoch": 2.1590128482381377, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.608548164367676, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8745306134223938, "num_tokens": 647516855.0, "step": 16972 }, { "epoch": 2.159140058516728, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.66545581817627, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8689640760421753, "num_tokens": 647556913.0, "step": 16973 }, { "epoch": 2.1592672687953187, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.636824607849121, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.8883281946182251, "num_tokens": 647591242.0, "step": 16974 }, { "epoch": 2.1593944790739092, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.59888744354248, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8765597343444824, "num_tokens": 647632156.0, "step": 16975 }, { "epoch": 2.1595216893524998, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.686726570129395, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8781132102012634, "num_tokens": 647669857.0, "step": 16976 }, { "epoch": 2.1596488996310903, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.577876091003418, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8604251742362976, "num_tokens": 647715580.0, "step": 16977 }, { "epoch": 2.159776109909681, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.656166076660156, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8637495040893555, "num_tokens": 647755917.0, "step": 16978 }, { "epoch": 2.1599033201882714, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.592896461486816, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8889931440353394, "num_tokens": 647802205.0, "step": 16979 }, { "epoch": 2.160030530466862, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.701948165893555, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8589537143707275, "num_tokens": 647846627.0, "step": 16980 }, { "epoch": 2.1601577407454524, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.583569526672363, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8640410900115967, "num_tokens": 647885033.0, "step": 16981 }, { "epoch": 2.160284951024043, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.636055946350098, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8631447553634644, "num_tokens": 647923370.0, "step": 16982 }, { "epoch": 2.1604121613026335, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.650226593017578, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8785675764083862, "num_tokens": 647963431.0, "step": 16983 }, { "epoch": 2.1605393715812236, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.582060813903809, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8645329475402832, "num_tokens": 648002854.0, "step": 16984 }, { "epoch": 2.160666581859814, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.616744995117188, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8605751395225525, "num_tokens": 648041082.0, "step": 16985 }, { "epoch": 2.1607937921384046, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.620142936706543, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8647428750991821, "num_tokens": 648082109.0, "step": 16986 }, { "epoch": 2.160921002416995, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.58251953125, "learning_rate": 1e-06, "loss": 0.348, "mean_token_accuracy": 0.8862807154655457, "num_tokens": 648117398.0, "step": 16987 }, { "epoch": 2.1610482126955857, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.55512523651123, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8794935345649719, "num_tokens": 648158997.0, "step": 16988 }, { "epoch": 2.161175422974176, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.638700485229492, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.870071530342102, "num_tokens": 648194334.0, "step": 16989 }, { "epoch": 2.1613026332527667, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.58434772491455, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.869570791721344, "num_tokens": 648230601.0, "step": 16990 }, { "epoch": 2.1614298435313573, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.642462730407715, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8667974472045898, "num_tokens": 648274979.0, "step": 16991 }, { "epoch": 2.161557053809948, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.58012866973877, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8569884300231934, "num_tokens": 648319229.0, "step": 16992 }, { "epoch": 2.1616842640885383, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.70505428314209, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8846220970153809, "num_tokens": 648354550.0, "step": 16993 }, { "epoch": 2.161811474367129, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.623963356018066, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8846420049667358, "num_tokens": 648389820.0, "step": 16994 }, { "epoch": 2.1619386846457194, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.704941749572754, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8866039514541626, "num_tokens": 648423841.0, "step": 16995 }, { "epoch": 2.16206589492431, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.634161949157715, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8597502708435059, "num_tokens": 648459865.0, "step": 16996 }, { "epoch": 2.1621931052029004, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.696364402770996, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.8898881673812866, "num_tokens": 648492639.0, "step": 16997 }, { "epoch": 2.162320315481491, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.605485916137695, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8775047063827515, "num_tokens": 648533456.0, "step": 16998 }, { "epoch": 2.1624475257600815, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.688457489013672, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8860810995101929, "num_tokens": 648571955.0, "step": 16999 }, { "epoch": 2.162574736038672, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.55781364440918, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.889220118522644, "num_tokens": 648614284.0, "step": 17000 }, { "epoch": 2.1627019463172625, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.612220764160156, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8808907270431519, "num_tokens": 648654399.0, "step": 17001 }, { "epoch": 2.162829156595853, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.587384223937988, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8753340244293213, "num_tokens": 648694490.0, "step": 17002 }, { "epoch": 2.1629563668744436, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.596701622009277, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8787819147109985, "num_tokens": 648730665.0, "step": 17003 }, { "epoch": 2.163083577153034, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.605295181274414, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8772982954978943, "num_tokens": 648765289.0, "step": 17004 }, { "epoch": 2.1632107874316246, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.605769157409668, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.873078465461731, "num_tokens": 648803982.0, "step": 17005 }, { "epoch": 2.163337997710215, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.584427833557129, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8783079385757446, "num_tokens": 648840992.0, "step": 17006 }, { "epoch": 2.1634652079888053, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.611531257629395, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8681610226631165, "num_tokens": 648883621.0, "step": 17007 }, { "epoch": 2.1635924182673962, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61004638671875, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8583678007125854, "num_tokens": 648922258.0, "step": 17008 }, { "epoch": 2.1637196285459863, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.54104995727539, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8720526099205017, "num_tokens": 648965010.0, "step": 17009 }, { "epoch": 2.163846838824577, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.589954376220703, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8676683902740479, "num_tokens": 649008124.0, "step": 17010 }, { "epoch": 2.1639740491031674, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.606075286865234, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8722639679908752, "num_tokens": 649044743.0, "step": 17011 }, { "epoch": 2.164101259381758, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.639370918273926, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8851478695869446, "num_tokens": 649078410.0, "step": 17012 }, { "epoch": 2.1642284696603484, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.60897445678711, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8834536671638489, "num_tokens": 649115415.0, "step": 17013 }, { "epoch": 2.164355679938939, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.652681350708008, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8702327013015747, "num_tokens": 649156266.0, "step": 17014 }, { "epoch": 2.1644828902175295, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.632425308227539, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.880344569683075, "num_tokens": 649192552.0, "step": 17015 }, { "epoch": 2.16461010049612, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.67193603515625, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.875575602054596, "num_tokens": 649224163.0, "step": 17016 }, { "epoch": 2.1647373107747105, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.610440254211426, "learning_rate": 1e-06, "loss": 0.3331, "mean_token_accuracy": 0.8918470144271851, "num_tokens": 649259348.0, "step": 17017 }, { "epoch": 2.164864521053301, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.667309761047363, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8887859582901001, "num_tokens": 649296875.0, "step": 17018 }, { "epoch": 2.1649917313318916, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.631617546081543, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8611844778060913, "num_tokens": 649333455.0, "step": 17019 }, { "epoch": 2.165118941610482, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.652320861816406, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8743438124656677, "num_tokens": 649364567.0, "step": 17020 }, { "epoch": 2.1652461518890727, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.694474220275879, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8559226989746094, "num_tokens": 649400579.0, "step": 17021 }, { "epoch": 2.165373362167663, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.593852043151855, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8850896954536438, "num_tokens": 649432940.0, "step": 17022 }, { "epoch": 2.1655005724462537, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.685340881347656, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8766003847122192, "num_tokens": 649466695.0, "step": 17023 }, { "epoch": 2.1656277827248442, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.584949493408203, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8717501759529114, "num_tokens": 649504578.0, "step": 17024 }, { "epoch": 2.1657549930034348, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.631460189819336, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8635335564613342, "num_tokens": 649541536.0, "step": 17025 }, { "epoch": 2.1658822032820253, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.62818431854248, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8715415000915527, "num_tokens": 649581280.0, "step": 17026 }, { "epoch": 2.166009413560616, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.654865264892578, "learning_rate": 1e-06, "loss": 0.3388, "mean_token_accuracy": 0.894644021987915, "num_tokens": 649619837.0, "step": 17027 }, { "epoch": 2.1661366238392064, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.657465934753418, "learning_rate": 1e-06, "loss": 0.3557, "mean_token_accuracy": 0.8844270706176758, "num_tokens": 649654708.0, "step": 17028 }, { "epoch": 2.166263834117797, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.67423152923584, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8528919219970703, "num_tokens": 649691709.0, "step": 17029 }, { "epoch": 2.1663910443963874, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.609596252441406, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8695554733276367, "num_tokens": 649734876.0, "step": 17030 }, { "epoch": 2.166518254674978, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.60617446899414, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8815464973449707, "num_tokens": 649775477.0, "step": 17031 }, { "epoch": 2.166645464953568, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.65749454498291, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8847355842590332, "num_tokens": 649816136.0, "step": 17032 }, { "epoch": 2.1667726752321586, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.647047996520996, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8823304772377014, "num_tokens": 649847942.0, "step": 17033 }, { "epoch": 2.166899885510749, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.595959663391113, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8701609373092651, "num_tokens": 649887875.0, "step": 17034 }, { "epoch": 2.1670270957893396, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.649723052978516, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8616615533828735, "num_tokens": 649920557.0, "step": 17035 }, { "epoch": 2.16715430606793, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.598417282104492, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8816480040550232, "num_tokens": 649959704.0, "step": 17036 }, { "epoch": 2.1672815163465207, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.590215682983398, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8779198527336121, "num_tokens": 650000433.0, "step": 17037 }, { "epoch": 2.167408726625111, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.592781066894531, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8859243988990784, "num_tokens": 650038793.0, "step": 17038 }, { "epoch": 2.1675359369037017, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.579258918762207, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8818480968475342, "num_tokens": 650072877.0, "step": 17039 }, { "epoch": 2.1676631471822922, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.632875442504883, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8609957695007324, "num_tokens": 650109245.0, "step": 17040 }, { "epoch": 2.1677903574608828, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.562045097351074, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.886293351650238, "num_tokens": 650150254.0, "step": 17041 }, { "epoch": 2.1679175677394733, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.635883331298828, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8548055291175842, "num_tokens": 650189587.0, "step": 17042 }, { "epoch": 2.168044778018064, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.600798606872559, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8694860339164734, "num_tokens": 650225443.0, "step": 17043 }, { "epoch": 2.1681719882966544, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.5908203125, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.869599461555481, "num_tokens": 650262319.0, "step": 17044 }, { "epoch": 2.168299198575245, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.582625389099121, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8584140539169312, "num_tokens": 650297137.0, "step": 17045 }, { "epoch": 2.1684264088538354, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.657247543334961, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8719264268875122, "num_tokens": 650335521.0, "step": 17046 }, { "epoch": 2.168553619132426, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.543554306030273, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8681287169456482, "num_tokens": 650373050.0, "step": 17047 }, { "epoch": 2.1686808294110165, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.631095886230469, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8798304796218872, "num_tokens": 650407438.0, "step": 17048 }, { "epoch": 2.168808039689607, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.597549438476562, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.867997407913208, "num_tokens": 650447793.0, "step": 17049 }, { "epoch": 2.1689352499681975, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.61469554901123, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8693714141845703, "num_tokens": 650487167.0, "step": 17050 }, { "epoch": 2.169062460246788, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.620786666870117, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8606879711151123, "num_tokens": 650522960.0, "step": 17051 }, { "epoch": 2.1691896705253786, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.570698738098145, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8673616647720337, "num_tokens": 650564161.0, "step": 17052 }, { "epoch": 2.169316880803969, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.658912658691406, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8793534636497498, "num_tokens": 650599201.0, "step": 17053 }, { "epoch": 2.1694440910825596, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.658263206481934, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8548049926757812, "num_tokens": 650637385.0, "step": 17054 }, { "epoch": 2.16957130136115, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.619878768920898, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8739370703697205, "num_tokens": 650671954.0, "step": 17055 }, { "epoch": 2.1696985116397407, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.642237663269043, "learning_rate": 1e-06, "loss": 0.344, "mean_token_accuracy": 0.8902721405029297, "num_tokens": 650708475.0, "step": 17056 }, { "epoch": 2.169825721918331, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.534151077270508, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8768497705459595, "num_tokens": 650748089.0, "step": 17057 }, { "epoch": 2.1699529321969213, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.614544868469238, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8755992650985718, "num_tokens": 650790869.0, "step": 17058 }, { "epoch": 2.170080142475512, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.627071380615234, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8607260584831238, "num_tokens": 650830697.0, "step": 17059 }, { "epoch": 2.1702073527541024, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.643306732177734, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8814616203308105, "num_tokens": 650868755.0, "step": 17060 }, { "epoch": 2.170334563032693, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.56173038482666, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8837827444076538, "num_tokens": 650908019.0, "step": 17061 }, { "epoch": 2.1704617733112834, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.603784561157227, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8717955946922302, "num_tokens": 650954195.0, "step": 17062 }, { "epoch": 2.170588983589874, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.597174644470215, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.886330783367157, "num_tokens": 650992740.0, "step": 17063 }, { "epoch": 2.1707161938684645, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.610730171203613, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8864735960960388, "num_tokens": 651031524.0, "step": 17064 }, { "epoch": 2.170843404147055, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.653579711914062, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8710744976997375, "num_tokens": 651070004.0, "step": 17065 }, { "epoch": 2.1709706144256455, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.589325904846191, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8599382638931274, "num_tokens": 651104474.0, "step": 17066 }, { "epoch": 2.171097824704236, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.497899055480957, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8669254183769226, "num_tokens": 651148913.0, "step": 17067 }, { "epoch": 2.1712250349828266, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61363410949707, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8813833594322205, "num_tokens": 651192072.0, "step": 17068 }, { "epoch": 2.171352245261417, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.552911758422852, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.865561306476593, "num_tokens": 651226132.0, "step": 17069 }, { "epoch": 2.1714794555400077, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.64745807647705, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8761274218559265, "num_tokens": 651263467.0, "step": 17070 }, { "epoch": 2.171606665818598, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.5323486328125, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8643313646316528, "num_tokens": 651309521.0, "step": 17071 }, { "epoch": 2.1717338760971887, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.623770713806152, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8756452798843384, "num_tokens": 651347026.0, "step": 17072 }, { "epoch": 2.1718610863757792, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.637887001037598, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8823930025100708, "num_tokens": 651380545.0, "step": 17073 }, { "epoch": 2.1719882966543698, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.710036277770996, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8779123425483704, "num_tokens": 651409428.0, "step": 17074 }, { "epoch": 2.1721155069329603, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.552749633789062, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8605546951293945, "num_tokens": 651447993.0, "step": 17075 }, { "epoch": 2.172242717211551, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.65860366821289, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8682518005371094, "num_tokens": 651488210.0, "step": 17076 }, { "epoch": 2.1723699274901414, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.529983520507812, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8690080642700195, "num_tokens": 651525742.0, "step": 17077 }, { "epoch": 2.172497137768732, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.669846534729004, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8800809383392334, "num_tokens": 651565569.0, "step": 17078 }, { "epoch": 2.1726243480473224, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.608965873718262, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8687658309936523, "num_tokens": 651605291.0, "step": 17079 }, { "epoch": 2.172751558325913, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.67700481414795, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8736768960952759, "num_tokens": 651641998.0, "step": 17080 }, { "epoch": 2.1728787686045035, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.618877410888672, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8706152439117432, "num_tokens": 651683040.0, "step": 17081 }, { "epoch": 2.1730059788830935, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.710602760314941, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8823056221008301, "num_tokens": 651718067.0, "step": 17082 }, { "epoch": 2.173133189161684, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.616021156311035, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8871729373931885, "num_tokens": 651761995.0, "step": 17083 }, { "epoch": 2.1732603994402746, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.630855560302734, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8864537477493286, "num_tokens": 651797780.0, "step": 17084 }, { "epoch": 2.173387609718865, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.587748527526855, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8627170324325562, "num_tokens": 651832931.0, "step": 17085 }, { "epoch": 2.1735148199974557, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.70046615600586, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8673866391181946, "num_tokens": 651868440.0, "step": 17086 }, { "epoch": 2.173642030276046, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.649802207946777, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8763773441314697, "num_tokens": 651900029.0, "step": 17087 }, { "epoch": 2.1737692405546367, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.636853218078613, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8699817657470703, "num_tokens": 651944892.0, "step": 17088 }, { "epoch": 2.1738964508332272, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.640069961547852, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8814929723739624, "num_tokens": 651979330.0, "step": 17089 }, { "epoch": 2.1740236611118178, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.642586708068848, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8649627566337585, "num_tokens": 652017389.0, "step": 17090 }, { "epoch": 2.1741508713904083, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.641762733459473, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8703010678291321, "num_tokens": 652052260.0, "step": 17091 }, { "epoch": 2.174278081668999, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.627776145935059, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8585806488990784, "num_tokens": 652089917.0, "step": 17092 }, { "epoch": 2.1744052919475894, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.630780220031738, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8760091066360474, "num_tokens": 652129849.0, "step": 17093 }, { "epoch": 2.17453250222618, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.646933555603027, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8657979369163513, "num_tokens": 652163178.0, "step": 17094 }, { "epoch": 2.1746597125047704, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.625104904174805, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8786600828170776, "num_tokens": 652201521.0, "step": 17095 }, { "epoch": 2.174786922783361, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.63459587097168, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8742965459823608, "num_tokens": 652234820.0, "step": 17096 }, { "epoch": 2.1749141330619515, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.580962181091309, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8693060278892517, "num_tokens": 652275686.0, "step": 17097 }, { "epoch": 2.175041343340542, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.575724601745605, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8647294044494629, "num_tokens": 652313243.0, "step": 17098 }, { "epoch": 2.1751685536191325, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.6527681350708, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8878687024116516, "num_tokens": 652351550.0, "step": 17099 }, { "epoch": 2.175295763897723, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.59857177734375, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8705470561981201, "num_tokens": 652399834.0, "step": 17100 }, { "epoch": 2.1754229741763136, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.664466857910156, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8596407175064087, "num_tokens": 652435839.0, "step": 17101 }, { "epoch": 2.175550184454904, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.575472831726074, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8691461086273193, "num_tokens": 652475747.0, "step": 17102 }, { "epoch": 2.1756773947334946, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.607208251953125, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8853620290756226, "num_tokens": 652517147.0, "step": 17103 }, { "epoch": 2.175804605012085, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.724729537963867, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8855210542678833, "num_tokens": 652552811.0, "step": 17104 }, { "epoch": 2.1759318152906753, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.56125545501709, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8811811208724976, "num_tokens": 652587709.0, "step": 17105 }, { "epoch": 2.1760590255692662, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.595210075378418, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8728388547897339, "num_tokens": 652626213.0, "step": 17106 }, { "epoch": 2.1761862358478563, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.604598045349121, "learning_rate": 1e-06, "loss": 0.4943, "mean_token_accuracy": 0.8442184925079346, "num_tokens": 652665692.0, "step": 17107 }, { "epoch": 2.176313446126447, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.623963356018066, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8655562996864319, "num_tokens": 652705515.0, "step": 17108 }, { "epoch": 2.1764406564050374, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.623830795288086, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8850711584091187, "num_tokens": 652743343.0, "step": 17109 }, { "epoch": 2.176567866683628, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.551732063293457, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8678176999092102, "num_tokens": 652784587.0, "step": 17110 }, { "epoch": 2.1766950769622184, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.687445640563965, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8649119138717651, "num_tokens": 652821985.0, "step": 17111 }, { "epoch": 2.176822287240809, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.64426040649414, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8832229375839233, "num_tokens": 652855772.0, "step": 17112 }, { "epoch": 2.1769494975193995, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.572235107421875, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8694763779640198, "num_tokens": 652890497.0, "step": 17113 }, { "epoch": 2.17707670779799, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.576733589172363, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8834827542304993, "num_tokens": 652931611.0, "step": 17114 }, { "epoch": 2.1772039180765805, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.586567878723145, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8739657402038574, "num_tokens": 652972129.0, "step": 17115 }, { "epoch": 2.177331128355171, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.59266471862793, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8666837811470032, "num_tokens": 653009848.0, "step": 17116 }, { "epoch": 2.1774583386337616, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.58690357208252, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8699419498443604, "num_tokens": 653044614.0, "step": 17117 }, { "epoch": 2.177585548912352, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.561482429504395, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8865638971328735, "num_tokens": 653087614.0, "step": 17118 }, { "epoch": 2.1777127591909426, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.618440628051758, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.885729968547821, "num_tokens": 653120434.0, "step": 17119 }, { "epoch": 2.177839969469533, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.625954627990723, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8833504915237427, "num_tokens": 653159390.0, "step": 17120 }, { "epoch": 2.1779671797481237, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.540162086486816, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8656889796257019, "num_tokens": 653201045.0, "step": 17121 }, { "epoch": 2.1780943900267142, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.624641418457031, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.87591153383255, "num_tokens": 653244968.0, "step": 17122 }, { "epoch": 2.1782216003053048, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.585156440734863, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8662070035934448, "num_tokens": 653281157.0, "step": 17123 }, { "epoch": 2.1783488105838953, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.658848762512207, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8863295912742615, "num_tokens": 653321055.0, "step": 17124 }, { "epoch": 2.178476020862486, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.574881553649902, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8598939180374146, "num_tokens": 653359320.0, "step": 17125 }, { "epoch": 2.1786032311410763, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.590392112731934, "learning_rate": 1e-06, "loss": 0.3328, "mean_token_accuracy": 0.8927536010742188, "num_tokens": 653404186.0, "step": 17126 }, { "epoch": 2.178730441419667, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.567473411560059, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8839967250823975, "num_tokens": 653438122.0, "step": 17127 }, { "epoch": 2.1788576516982574, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.71377182006836, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.872320830821991, "num_tokens": 653474727.0, "step": 17128 }, { "epoch": 2.178984861976848, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.53602123260498, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8785444498062134, "num_tokens": 653513531.0, "step": 17129 }, { "epoch": 2.179112072255438, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.61677360534668, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8845110535621643, "num_tokens": 653555871.0, "step": 17130 }, { "epoch": 2.1792392825340285, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.598223686218262, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8752413988113403, "num_tokens": 653598825.0, "step": 17131 }, { "epoch": 2.179366492812619, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.626778602600098, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8824710845947266, "num_tokens": 653639838.0, "step": 17132 }, { "epoch": 2.1794937030912096, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.62721061706543, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8648152351379395, "num_tokens": 653678958.0, "step": 17133 }, { "epoch": 2.1796209133698, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.657782554626465, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8820285201072693, "num_tokens": 653718043.0, "step": 17134 }, { "epoch": 2.1797481236483907, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.602294921875, "learning_rate": 1e-06, "loss": 0.3509, "mean_token_accuracy": 0.8879134654998779, "num_tokens": 653756308.0, "step": 17135 }, { "epoch": 2.179875333926981, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.747138977050781, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8822116255760193, "num_tokens": 653793140.0, "step": 17136 }, { "epoch": 2.1800025442055717, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.641924858093262, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.879165530204773, "num_tokens": 653823441.0, "step": 17137 }, { "epoch": 2.1801297544841622, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.614043235778809, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8645672798156738, "num_tokens": 653860743.0, "step": 17138 }, { "epoch": 2.1802569647627528, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.601982116699219, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.886396050453186, "num_tokens": 653894785.0, "step": 17139 }, { "epoch": 2.1803841750413433, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.628480911254883, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8704924583435059, "num_tokens": 653932208.0, "step": 17140 }, { "epoch": 2.180511385319934, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.642861366271973, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8601369261741638, "num_tokens": 653968818.0, "step": 17141 }, { "epoch": 2.1806385955985244, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.499723434448242, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8692521452903748, "num_tokens": 654010696.0, "step": 17142 }, { "epoch": 2.180765805877115, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.637372970581055, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8779616355895996, "num_tokens": 654047632.0, "step": 17143 }, { "epoch": 2.1808930161557054, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.58857536315918, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.877464234828949, "num_tokens": 654088200.0, "step": 17144 }, { "epoch": 2.181020226434296, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.652575492858887, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8709407448768616, "num_tokens": 654124225.0, "step": 17145 }, { "epoch": 2.1811474367128865, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.62219524383545, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8759227991104126, "num_tokens": 654163317.0, "step": 17146 }, { "epoch": 2.181274646991477, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.619705200195312, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8658809661865234, "num_tokens": 654199208.0, "step": 17147 }, { "epoch": 2.1814018572700675, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.643588066101074, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8624904751777649, "num_tokens": 654239988.0, "step": 17148 }, { "epoch": 2.181529067548658, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.619725227355957, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8685092926025391, "num_tokens": 654276361.0, "step": 17149 }, { "epoch": 2.1816562778272486, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.6001615524292, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8766400814056396, "num_tokens": 654316690.0, "step": 17150 }, { "epoch": 2.181783488105839, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.596871376037598, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8831839561462402, "num_tokens": 654354112.0, "step": 17151 }, { "epoch": 2.1819106983844296, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.612517356872559, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.885168194770813, "num_tokens": 654398896.0, "step": 17152 }, { "epoch": 2.18203790866302, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.598673820495605, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8855706453323364, "num_tokens": 654435180.0, "step": 17153 }, { "epoch": 2.1821651189416107, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.69274616241455, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.862236738204956, "num_tokens": 654473958.0, "step": 17154 }, { "epoch": 2.1822923292202008, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.5816068649292, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8600972890853882, "num_tokens": 654512918.0, "step": 17155 }, { "epoch": 2.1824195394987913, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.70688247680664, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8811162710189819, "num_tokens": 654552296.0, "step": 17156 }, { "epoch": 2.182546749777382, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.631924629211426, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8811907768249512, "num_tokens": 654592088.0, "step": 17157 }, { "epoch": 2.1826739600559724, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.621502876281738, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8663464784622192, "num_tokens": 654635119.0, "step": 17158 }, { "epoch": 2.182801170334563, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.685245513916016, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8618183135986328, "num_tokens": 654671204.0, "step": 17159 }, { "epoch": 2.1829283806131534, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.571661949157715, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8728863000869751, "num_tokens": 654702986.0, "step": 17160 }, { "epoch": 2.183055590891744, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.609745025634766, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8503950834274292, "num_tokens": 654745813.0, "step": 17161 }, { "epoch": 2.1831828011703345, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.604205131530762, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8738600611686707, "num_tokens": 654783305.0, "step": 17162 }, { "epoch": 2.183310011448925, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.609612464904785, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8823937773704529, "num_tokens": 654823427.0, "step": 17163 }, { "epoch": 2.1834372217275155, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.607725143432617, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8863370418548584, "num_tokens": 654862400.0, "step": 17164 }, { "epoch": 2.183564432006106, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.545760154724121, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.878463864326477, "num_tokens": 654902484.0, "step": 17165 }, { "epoch": 2.1836916422846966, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.581323623657227, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8803820610046387, "num_tokens": 654946882.0, "step": 17166 }, { "epoch": 2.183818852563287, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.617823600769043, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8651636838912964, "num_tokens": 654977850.0, "step": 17167 }, { "epoch": 2.1839460628418776, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.605300903320312, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8600499629974365, "num_tokens": 655015352.0, "step": 17168 }, { "epoch": 2.184073273120468, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.572872161865234, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.874313473701477, "num_tokens": 655047488.0, "step": 17169 }, { "epoch": 2.1842004833990587, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.625908851623535, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8717098832130432, "num_tokens": 655083410.0, "step": 17170 }, { "epoch": 2.1843276936776492, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.548579216003418, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.882354199886322, "num_tokens": 655120594.0, "step": 17171 }, { "epoch": 2.1844549039562398, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.67663860321045, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8722856640815735, "num_tokens": 655157310.0, "step": 17172 }, { "epoch": 2.1845821142348303, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.562582015991211, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.883467972278595, "num_tokens": 655193749.0, "step": 17173 }, { "epoch": 2.184709324513421, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.6963529586792, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8880923986434937, "num_tokens": 655235532.0, "step": 17174 }, { "epoch": 2.1848365347920113, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.528265953063965, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8796644806861877, "num_tokens": 655282154.0, "step": 17175 }, { "epoch": 2.184963745070602, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.604230880737305, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8681359887123108, "num_tokens": 655321342.0, "step": 17176 }, { "epoch": 2.1850909553491924, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.561636924743652, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8910267949104309, "num_tokens": 655358107.0, "step": 17177 }, { "epoch": 2.185218165627783, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.639559745788574, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8694088459014893, "num_tokens": 655393462.0, "step": 17178 }, { "epoch": 2.1853453759063735, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.578662872314453, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8753719329833984, "num_tokens": 655436114.0, "step": 17179 }, { "epoch": 2.1854725861849635, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.593873023986816, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8757873773574829, "num_tokens": 655473119.0, "step": 17180 }, { "epoch": 2.185599796463554, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.599459648132324, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8690149784088135, "num_tokens": 655515537.0, "step": 17181 }, { "epoch": 2.1857270067421446, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.613088607788086, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8681076765060425, "num_tokens": 655555944.0, "step": 17182 }, { "epoch": 2.185854217020735, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.576712608337402, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8582039475440979, "num_tokens": 655589952.0, "step": 17183 }, { "epoch": 2.1859814272993257, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.546689987182617, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8860329389572144, "num_tokens": 655628269.0, "step": 17184 }, { "epoch": 2.186108637577916, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.584025382995605, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8765398263931274, "num_tokens": 655657315.0, "step": 17185 }, { "epoch": 2.1862358478565067, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.581978797912598, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8736273050308228, "num_tokens": 655696117.0, "step": 17186 }, { "epoch": 2.1863630581350972, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.642699241638184, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8568114042282104, "num_tokens": 655738823.0, "step": 17187 }, { "epoch": 2.1864902684136878, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.5347261428833, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8654429912567139, "num_tokens": 655781457.0, "step": 17188 }, { "epoch": 2.1866174786922783, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.632402420043945, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8685021996498108, "num_tokens": 655824704.0, "step": 17189 }, { "epoch": 2.186744688970869, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.591267585754395, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8828989267349243, "num_tokens": 655860013.0, "step": 17190 }, { "epoch": 2.1868718992494594, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.641416549682617, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8718609809875488, "num_tokens": 655897146.0, "step": 17191 }, { "epoch": 2.18699910952805, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.61920166015625, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8788626194000244, "num_tokens": 655934299.0, "step": 17192 }, { "epoch": 2.1871263198066404, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.549882888793945, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8769755363464355, "num_tokens": 655976554.0, "step": 17193 }, { "epoch": 2.187253530085231, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.594207763671875, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8851311206817627, "num_tokens": 656011756.0, "step": 17194 }, { "epoch": 2.1873807403638215, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.608059883117676, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8706487417221069, "num_tokens": 656054405.0, "step": 17195 }, { "epoch": 2.187507950642412, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.559199333190918, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8778380155563354, "num_tokens": 656094668.0, "step": 17196 }, { "epoch": 2.1876351609210025, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.54958724975586, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8739004135131836, "num_tokens": 656137664.0, "step": 17197 }, { "epoch": 2.187762371199593, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.650428771972656, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8665767312049866, "num_tokens": 656180611.0, "step": 17198 }, { "epoch": 2.1878895814781836, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.616728782653809, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8657334446907043, "num_tokens": 656217789.0, "step": 17199 }, { "epoch": 2.188016791756774, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.57238483428955, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8708405494689941, "num_tokens": 656261940.0, "step": 17200 }, { "epoch": 2.1881440020353646, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.580175399780273, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.883544385433197, "num_tokens": 656296000.0, "step": 17201 }, { "epoch": 2.188271212313955, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.635383605957031, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8838933706283569, "num_tokens": 656340862.0, "step": 17202 }, { "epoch": 2.1883984225925452, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.64450454711914, "learning_rate": 1e-06, "loss": 0.3264, "mean_token_accuracy": 0.8973563313484192, "num_tokens": 656376923.0, "step": 17203 }, { "epoch": 2.188525632871136, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.671951293945312, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.870811402797699, "num_tokens": 656414058.0, "step": 17204 }, { "epoch": 2.1886528431497263, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.560864448547363, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8654187321662903, "num_tokens": 656450077.0, "step": 17205 }, { "epoch": 2.188780053428317, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.726258277893066, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.865725576877594, "num_tokens": 656487258.0, "step": 17206 }, { "epoch": 2.1889072637069074, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.61233139038086, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8883692026138306, "num_tokens": 656525141.0, "step": 17207 }, { "epoch": 2.189034473985498, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.618768692016602, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8772967457771301, "num_tokens": 656561517.0, "step": 17208 }, { "epoch": 2.1891616842640884, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.61897087097168, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.879966139793396, "num_tokens": 656595859.0, "step": 17209 }, { "epoch": 2.189288894542679, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.62427806854248, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8703145384788513, "num_tokens": 656627652.0, "step": 17210 }, { "epoch": 2.1894161048212695, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.550070762634277, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8807278871536255, "num_tokens": 656669951.0, "step": 17211 }, { "epoch": 2.18954331509986, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.578560829162598, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8853857517242432, "num_tokens": 656702216.0, "step": 17212 }, { "epoch": 2.1896705253784505, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.581648826599121, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8638389706611633, "num_tokens": 656744816.0, "step": 17213 }, { "epoch": 2.189797735657041, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.581149101257324, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8700859546661377, "num_tokens": 656785533.0, "step": 17214 }, { "epoch": 2.1899249459356316, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.623250961303711, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8716127872467041, "num_tokens": 656825375.0, "step": 17215 }, { "epoch": 2.190052156214222, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.6113920211792, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8818007111549377, "num_tokens": 656857951.0, "step": 17216 }, { "epoch": 2.1901793664928126, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.645724296569824, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8831610679626465, "num_tokens": 656896911.0, "step": 17217 }, { "epoch": 2.190306576771403, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.636999130249023, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8810871839523315, "num_tokens": 656935723.0, "step": 17218 }, { "epoch": 2.1904337870499937, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.59334659576416, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8848788738250732, "num_tokens": 656970881.0, "step": 17219 }, { "epoch": 2.1905609973285842, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.590518951416016, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8694180846214294, "num_tokens": 657007721.0, "step": 17220 }, { "epoch": 2.1906882076071748, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.635544776916504, "learning_rate": 1e-06, "loss": 0.3466, "mean_token_accuracy": 0.8844236135482788, "num_tokens": 657042359.0, "step": 17221 }, { "epoch": 2.1908154178857653, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.575699806213379, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8799764513969421, "num_tokens": 657084662.0, "step": 17222 }, { "epoch": 2.190942628164356, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.692736625671387, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.875682532787323, "num_tokens": 657122002.0, "step": 17223 }, { "epoch": 2.1910698384429463, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.644505500793457, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8756018877029419, "num_tokens": 657158768.0, "step": 17224 }, { "epoch": 2.191197048721537, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.640073776245117, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8620964288711548, "num_tokens": 657193445.0, "step": 17225 }, { "epoch": 2.1913242590001274, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.627711296081543, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8730518817901611, "num_tokens": 657234922.0, "step": 17226 }, { "epoch": 2.191451469278718, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.639533042907715, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8653818368911743, "num_tokens": 657270774.0, "step": 17227 }, { "epoch": 2.191578679557308, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.665465354919434, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.888361930847168, "num_tokens": 657307046.0, "step": 17228 }, { "epoch": 2.1917058898358985, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.589534759521484, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8644665479660034, "num_tokens": 657352865.0, "step": 17229 }, { "epoch": 2.191833100114489, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.663982391357422, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8544611930847168, "num_tokens": 657393372.0, "step": 17230 }, { "epoch": 2.1919603103930796, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.657021522521973, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8536185026168823, "num_tokens": 657432445.0, "step": 17231 }, { "epoch": 2.19208752067167, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.673701286315918, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8719580173492432, "num_tokens": 657469145.0, "step": 17232 }, { "epoch": 2.1922147309502606, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.601812362670898, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8879669904708862, "num_tokens": 657510369.0, "step": 17233 }, { "epoch": 2.192341941228851, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.68693733215332, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8803865909576416, "num_tokens": 657550257.0, "step": 17234 }, { "epoch": 2.1924691515074417, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.640839576721191, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8764876127243042, "num_tokens": 657587095.0, "step": 17235 }, { "epoch": 2.1925963617860322, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.65468692779541, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8782729506492615, "num_tokens": 657619710.0, "step": 17236 }, { "epoch": 2.1927235720646228, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.719818115234375, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8698490858078003, "num_tokens": 657653316.0, "step": 17237 }, { "epoch": 2.1928507823432133, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.55955696105957, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8700472116470337, "num_tokens": 657693570.0, "step": 17238 }, { "epoch": 2.192977992621804, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.669244766235352, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8748038411140442, "num_tokens": 657736781.0, "step": 17239 }, { "epoch": 2.1931052029003943, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.692570686340332, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8661679029464722, "num_tokens": 657775603.0, "step": 17240 }, { "epoch": 2.193232413178985, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.614461898803711, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8718389868736267, "num_tokens": 657809065.0, "step": 17241 }, { "epoch": 2.1933596234575754, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.67656421661377, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8718501925468445, "num_tokens": 657851369.0, "step": 17242 }, { "epoch": 2.193486833736166, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.584091186523438, "learning_rate": 1e-06, "loss": 0.3372, "mean_token_accuracy": 0.8946504592895508, "num_tokens": 657889008.0, "step": 17243 }, { "epoch": 2.1936140440147565, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.666935920715332, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8722438812255859, "num_tokens": 657925867.0, "step": 17244 }, { "epoch": 2.193741254293347, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.69063663482666, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8642380237579346, "num_tokens": 657959502.0, "step": 17245 }, { "epoch": 2.1938684645719375, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.669547080993652, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8776410818099976, "num_tokens": 657999063.0, "step": 17246 }, { "epoch": 2.193995674850528, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.645339965820312, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8640893697738647, "num_tokens": 658033016.0, "step": 17247 }, { "epoch": 2.1941228851291186, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.70632266998291, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8809512853622437, "num_tokens": 658071259.0, "step": 17248 }, { "epoch": 2.194250095407709, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.699012756347656, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8754807710647583, "num_tokens": 658103190.0, "step": 17249 }, { "epoch": 2.1943773056862996, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.627952575683594, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8671913743019104, "num_tokens": 658143346.0, "step": 17250 }, { "epoch": 2.19450451596489, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.600247383117676, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.883155107498169, "num_tokens": 658181747.0, "step": 17251 }, { "epoch": 2.1946317262434807, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.67210578918457, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8762727379798889, "num_tokens": 658221770.0, "step": 17252 }, { "epoch": 2.1947589365220708, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.636824607849121, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8785520195960999, "num_tokens": 658260157.0, "step": 17253 }, { "epoch": 2.1948861468006613, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.610751152038574, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8640902042388916, "num_tokens": 658301815.0, "step": 17254 }, { "epoch": 2.195013357079252, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.65135669708252, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8762919902801514, "num_tokens": 658340732.0, "step": 17255 }, { "epoch": 2.1951405673578424, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.619911193847656, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8687359094619751, "num_tokens": 658380569.0, "step": 17256 }, { "epoch": 2.195267777636433, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.668830871582031, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8857447504997253, "num_tokens": 658415842.0, "step": 17257 }, { "epoch": 2.1953949879150234, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.626419067382812, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8779392242431641, "num_tokens": 658452861.0, "step": 17258 }, { "epoch": 2.195522198193614, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.630346298217773, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8769890069961548, "num_tokens": 658487019.0, "step": 17259 }, { "epoch": 2.1956494084722045, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.678622245788574, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8728245496749878, "num_tokens": 658529290.0, "step": 17260 }, { "epoch": 2.195776618750795, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.637331008911133, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8694794774055481, "num_tokens": 658572118.0, "step": 17261 }, { "epoch": 2.1959038290293855, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.623998641967773, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8710237741470337, "num_tokens": 658615972.0, "step": 17262 }, { "epoch": 2.196031039307976, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.648482322692871, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8815234899520874, "num_tokens": 658653848.0, "step": 17263 }, { "epoch": 2.1961582495865666, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.650290489196777, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8737831115722656, "num_tokens": 658686247.0, "step": 17264 }, { "epoch": 2.196285459865157, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.665828704833984, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8716110587120056, "num_tokens": 658725787.0, "step": 17265 }, { "epoch": 2.1964126701437476, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.643471717834473, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8703833818435669, "num_tokens": 658765957.0, "step": 17266 }, { "epoch": 2.196539880422338, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.589896202087402, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8841896057128906, "num_tokens": 658800996.0, "step": 17267 }, { "epoch": 2.1966670907009287, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.656007766723633, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8714777231216431, "num_tokens": 658833779.0, "step": 17268 }, { "epoch": 2.196794300979519, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.579845428466797, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8773160576820374, "num_tokens": 658874425.0, "step": 17269 }, { "epoch": 2.1969215112581097, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.667020797729492, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.882584810256958, "num_tokens": 658912007.0, "step": 17270 }, { "epoch": 2.1970487215367003, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.591050148010254, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8654972314834595, "num_tokens": 658956775.0, "step": 17271 }, { "epoch": 2.197175931815291, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.655755043029785, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8864666223526001, "num_tokens": 658997132.0, "step": 17272 }, { "epoch": 2.1973031420938813, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.640671730041504, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8810834288597107, "num_tokens": 659040437.0, "step": 17273 }, { "epoch": 2.197430352372472, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.664462089538574, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8891093730926514, "num_tokens": 659074461.0, "step": 17274 }, { "epoch": 2.1975575626510624, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.63059139251709, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8717342019081116, "num_tokens": 659113818.0, "step": 17275 }, { "epoch": 2.197684772929653, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.607444763183594, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8616799116134644, "num_tokens": 659149297.0, "step": 17276 }, { "epoch": 2.1978119832082434, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.68785572052002, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8773485422134399, "num_tokens": 659187500.0, "step": 17277 }, { "epoch": 2.1979391934868335, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.612166404724121, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8483387231826782, "num_tokens": 659228624.0, "step": 17278 }, { "epoch": 2.198066403765424, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.576393127441406, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8815709948539734, "num_tokens": 659270317.0, "step": 17279 }, { "epoch": 2.1981936140440146, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.666059494018555, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8744622468948364, "num_tokens": 659302133.0, "step": 17280 }, { "epoch": 2.198320824322605, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.539731979370117, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8800309896469116, "num_tokens": 659342469.0, "step": 17281 }, { "epoch": 2.1984480346011956, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.592267990112305, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8853170871734619, "num_tokens": 659373786.0, "step": 17282 }, { "epoch": 2.198575244879786, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.615622520446777, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8677127957344055, "num_tokens": 659412498.0, "step": 17283 }, { "epoch": 2.1987024551583767, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.606904983520508, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.859147310256958, "num_tokens": 659446415.0, "step": 17284 }, { "epoch": 2.1988296654369672, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.628463745117188, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8659307956695557, "num_tokens": 659488824.0, "step": 17285 }, { "epoch": 2.1989568757155578, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.647212982177734, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8780139684677124, "num_tokens": 659523036.0, "step": 17286 }, { "epoch": 2.1990840859941483, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.56756591796875, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8679588437080383, "num_tokens": 659562780.0, "step": 17287 }, { "epoch": 2.199211296272739, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.622495651245117, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8808273077011108, "num_tokens": 659599364.0, "step": 17288 }, { "epoch": 2.1993385065513293, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.621447563171387, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8683651685714722, "num_tokens": 659636275.0, "step": 17289 }, { "epoch": 2.19946571682992, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.580979347229004, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8739073872566223, "num_tokens": 659670315.0, "step": 17290 }, { "epoch": 2.1995929271085104, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.641670227050781, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8696503639221191, "num_tokens": 659710968.0, "step": 17291 }, { "epoch": 2.199720137387101, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.521978378295898, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8829299807548523, "num_tokens": 659751334.0, "step": 17292 }, { "epoch": 2.1998473476656915, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.6149263381958, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8743395805358887, "num_tokens": 659793867.0, "step": 17293 }, { "epoch": 2.199974557944282, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.551594734191895, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8744841814041138, "num_tokens": 659838724.0, "step": 17294 }, { "epoch": 2.2001017682228725, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.682638168334961, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8725504279136658, "num_tokens": 659876589.0, "step": 17295 }, { "epoch": 2.200228978501463, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.533370018005371, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8645834922790527, "num_tokens": 659914864.0, "step": 17296 }, { "epoch": 2.2003561887800536, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.622446060180664, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8801531791687012, "num_tokens": 659952482.0, "step": 17297 }, { "epoch": 2.200483399058644, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.60606575012207, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8746374845504761, "num_tokens": 659985192.0, "step": 17298 }, { "epoch": 2.2006106093372346, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.62545394897461, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8819417357444763, "num_tokens": 660019505.0, "step": 17299 }, { "epoch": 2.200737819615825, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.629528045654297, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8674614429473877, "num_tokens": 660054942.0, "step": 17300 }, { "epoch": 2.2008650298944152, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.566699981689453, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8739490509033203, "num_tokens": 660092836.0, "step": 17301 }, { "epoch": 2.200992240173006, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.636187553405762, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8693093061447144, "num_tokens": 660129419.0, "step": 17302 }, { "epoch": 2.2011194504515963, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.62627124786377, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.877509593963623, "num_tokens": 660168116.0, "step": 17303 }, { "epoch": 2.201246660730187, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.603979110717773, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8678755760192871, "num_tokens": 660208006.0, "step": 17304 }, { "epoch": 2.2013738710087773, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.600486755371094, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8733595609664917, "num_tokens": 660254046.0, "step": 17305 }, { "epoch": 2.201501081287368, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.606490135192871, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8722995519638062, "num_tokens": 660289321.0, "step": 17306 }, { "epoch": 2.2016282915659584, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.615540504455566, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.871177613735199, "num_tokens": 660325052.0, "step": 17307 }, { "epoch": 2.201755501844549, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.664449691772461, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8655465841293335, "num_tokens": 660357942.0, "step": 17308 }, { "epoch": 2.2018827121231395, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.61697006225586, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8486734628677368, "num_tokens": 660390654.0, "step": 17309 }, { "epoch": 2.20200992240173, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.583724021911621, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8747626543045044, "num_tokens": 660432020.0, "step": 17310 }, { "epoch": 2.2021371326803205, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.623064041137695, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.87433260679245, "num_tokens": 660465051.0, "step": 17311 }, { "epoch": 2.202264342958911, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.626472473144531, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8703656196594238, "num_tokens": 660504967.0, "step": 17312 }, { "epoch": 2.2023915532375016, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.589542388916016, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8644724488258362, "num_tokens": 660543826.0, "step": 17313 }, { "epoch": 2.202518763516092, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.619367599487305, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8697270154953003, "num_tokens": 660583074.0, "step": 17314 }, { "epoch": 2.2026459737946826, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.669147491455078, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8896629810333252, "num_tokens": 660619974.0, "step": 17315 }, { "epoch": 2.202773184073273, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.638362884521484, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8800156116485596, "num_tokens": 660654005.0, "step": 17316 }, { "epoch": 2.2029003943518637, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.666879653930664, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8624604940414429, "num_tokens": 660689179.0, "step": 17317 }, { "epoch": 2.203027604630454, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.68481731414795, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.862555205821991, "num_tokens": 660724923.0, "step": 17318 }, { "epoch": 2.2031548149090447, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.692839622497559, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8778834342956543, "num_tokens": 660760763.0, "step": 17319 }, { "epoch": 2.2032820251876353, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.638656616210938, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8738757371902466, "num_tokens": 660795520.0, "step": 17320 }, { "epoch": 2.203409235466226, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.644516944885254, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8839234709739685, "num_tokens": 660834858.0, "step": 17321 }, { "epoch": 2.2035364457448163, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.636096954345703, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8815184831619263, "num_tokens": 660870319.0, "step": 17322 }, { "epoch": 2.203663656023407, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.667312622070312, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8606471419334412, "num_tokens": 660907313.0, "step": 17323 }, { "epoch": 2.2037908663019974, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.584352493286133, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8884097337722778, "num_tokens": 660948087.0, "step": 17324 }, { "epoch": 2.203918076580588, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.602818489074707, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.879467248916626, "num_tokens": 660984984.0, "step": 17325 }, { "epoch": 2.204045286859178, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.698857307434082, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8713356852531433, "num_tokens": 661022898.0, "step": 17326 }, { "epoch": 2.2041724971377685, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.558622360229492, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8829710483551025, "num_tokens": 661058645.0, "step": 17327 }, { "epoch": 2.204299707416359, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.639946937561035, "learning_rate": 1e-06, "loss": 0.3382, "mean_token_accuracy": 0.8969086408615112, "num_tokens": 661096954.0, "step": 17328 }, { "epoch": 2.2044269176949496, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.658443450927734, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8637484908103943, "num_tokens": 661136947.0, "step": 17329 }, { "epoch": 2.20455412797354, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.705565452575684, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8759557008743286, "num_tokens": 661175376.0, "step": 17330 }, { "epoch": 2.2046813382521306, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.65164566040039, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8668996095657349, "num_tokens": 661213238.0, "step": 17331 }, { "epoch": 2.204808548530721, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.721059799194336, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8895819783210754, "num_tokens": 661246512.0, "step": 17332 }, { "epoch": 2.2049357588093117, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.642955780029297, "learning_rate": 1e-06, "loss": 0.3595, "mean_token_accuracy": 0.8853261470794678, "num_tokens": 661288272.0, "step": 17333 }, { "epoch": 2.2050629690879022, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.701144218444824, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8794302940368652, "num_tokens": 661328705.0, "step": 17334 }, { "epoch": 2.2051901793664928, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.640050888061523, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.878239631652832, "num_tokens": 661369877.0, "step": 17335 }, { "epoch": 2.2053173896450833, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.702701568603516, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8768705129623413, "num_tokens": 661405209.0, "step": 17336 }, { "epoch": 2.205444599923674, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.682709693908691, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8780519962310791, "num_tokens": 661441083.0, "step": 17337 }, { "epoch": 2.2055718102022643, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.696535110473633, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8772302269935608, "num_tokens": 661476938.0, "step": 17338 }, { "epoch": 2.205699020480855, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.715810775756836, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8663613796234131, "num_tokens": 661519281.0, "step": 17339 }, { "epoch": 2.2058262307594454, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.636441230773926, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.8895280361175537, "num_tokens": 661555880.0, "step": 17340 }, { "epoch": 2.205953441038036, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.666892051696777, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8661874532699585, "num_tokens": 661595937.0, "step": 17341 }, { "epoch": 2.2060806513166265, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.627903938293457, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8612350821495056, "num_tokens": 661636868.0, "step": 17342 }, { "epoch": 2.206207861595217, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.737714767456055, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8674710988998413, "num_tokens": 661681503.0, "step": 17343 }, { "epoch": 2.2063350718738075, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.684929847717285, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8790494203567505, "num_tokens": 661718438.0, "step": 17344 }, { "epoch": 2.206462282152398, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.687638282775879, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8545400500297546, "num_tokens": 661759048.0, "step": 17345 }, { "epoch": 2.2065894924309886, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 14.201971054077148, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.883978009223938, "num_tokens": 661803131.0, "step": 17346 }, { "epoch": 2.206716702709579, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.623961448669434, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8750413656234741, "num_tokens": 661838139.0, "step": 17347 }, { "epoch": 2.2068439129881696, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.635539054870605, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8682126998901367, "num_tokens": 661878568.0, "step": 17348 }, { "epoch": 2.20697112326676, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.725929260253906, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8684239387512207, "num_tokens": 661909110.0, "step": 17349 }, { "epoch": 2.2070983335453507, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.625049591064453, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8873242735862732, "num_tokens": 661944472.0, "step": 17350 }, { "epoch": 2.2072255438239408, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.718172073364258, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8859453797340393, "num_tokens": 661985865.0, "step": 17351 }, { "epoch": 2.2073527541025313, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.683209419250488, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8691934943199158, "num_tokens": 662019460.0, "step": 17352 }, { "epoch": 2.207479964381122, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.682998657226562, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.881356954574585, "num_tokens": 662059109.0, "step": 17353 }, { "epoch": 2.2076071746597123, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.639213562011719, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8674736022949219, "num_tokens": 662096040.0, "step": 17354 }, { "epoch": 2.207734384938303, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.691169738769531, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8790473937988281, "num_tokens": 662139351.0, "step": 17355 }, { "epoch": 2.2078615952168934, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.624951362609863, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8750524520874023, "num_tokens": 662177353.0, "step": 17356 }, { "epoch": 2.207988805495484, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.711384773254395, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8644007444381714, "num_tokens": 662209519.0, "step": 17357 }, { "epoch": 2.2081160157740745, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.694007873535156, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8557873368263245, "num_tokens": 662251311.0, "step": 17358 }, { "epoch": 2.208243226052665, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.739164352416992, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8816695213317871, "num_tokens": 662282770.0, "step": 17359 }, { "epoch": 2.2083704363312555, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.679780960083008, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8707448840141296, "num_tokens": 662326770.0, "step": 17360 }, { "epoch": 2.208497646609846, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.700547218322754, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.874235987663269, "num_tokens": 662366811.0, "step": 17361 }, { "epoch": 2.2086248568884366, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.714083671569824, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8699342012405396, "num_tokens": 662404517.0, "step": 17362 }, { "epoch": 2.208752067167027, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.686930656433105, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8800745010375977, "num_tokens": 662442091.0, "step": 17363 }, { "epoch": 2.2088792774456176, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.63815689086914, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8490145206451416, "num_tokens": 662483343.0, "step": 17364 }, { "epoch": 2.209006487724208, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.654170036315918, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8680691123008728, "num_tokens": 662521994.0, "step": 17365 }, { "epoch": 2.2091336980027987, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.701922416687012, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8696043491363525, "num_tokens": 662562737.0, "step": 17366 }, { "epoch": 2.209260908281389, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.675335884094238, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.870776891708374, "num_tokens": 662596927.0, "step": 17367 }, { "epoch": 2.2093881185599797, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.687322616577148, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8825832605361938, "num_tokens": 662632831.0, "step": 17368 }, { "epoch": 2.2095153288385703, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.782350540161133, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8711481094360352, "num_tokens": 662675554.0, "step": 17369 }, { "epoch": 2.209642539117161, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.614105224609375, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.885429859161377, "num_tokens": 662720248.0, "step": 17370 }, { "epoch": 2.2097697493957513, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.598502159118652, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8738387823104858, "num_tokens": 662761388.0, "step": 17371 }, { "epoch": 2.209896959674342, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.725556373596191, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8731701374053955, "num_tokens": 662798007.0, "step": 17372 }, { "epoch": 2.2100241699529324, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.667867660522461, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8710188865661621, "num_tokens": 662836042.0, "step": 17373 }, { "epoch": 2.210151380231523, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.654221534729004, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8644475936889648, "num_tokens": 662869465.0, "step": 17374 }, { "epoch": 2.2102785905101134, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.676276206970215, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8658449053764343, "num_tokens": 662909440.0, "step": 17375 }, { "epoch": 2.2104058007887035, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.63988208770752, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8699305057525635, "num_tokens": 662949637.0, "step": 17376 }, { "epoch": 2.210533011067294, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.63459587097168, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8720216751098633, "num_tokens": 662987679.0, "step": 17377 }, { "epoch": 2.2106602213458846, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.701566696166992, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.854917585849762, "num_tokens": 663027651.0, "step": 17378 }, { "epoch": 2.210787431624475, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.67900276184082, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8870362043380737, "num_tokens": 663063708.0, "step": 17379 }, { "epoch": 2.2109146419030656, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.692299842834473, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8658470511436462, "num_tokens": 663097255.0, "step": 17380 }, { "epoch": 2.211041852181656, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.64122200012207, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8704252243041992, "num_tokens": 663134496.0, "step": 17381 }, { "epoch": 2.2111690624602467, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.727910041809082, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8821619749069214, "num_tokens": 663170278.0, "step": 17382 }, { "epoch": 2.211296272738837, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.599884033203125, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.870997428894043, "num_tokens": 663205407.0, "step": 17383 }, { "epoch": 2.2114234830174277, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.722291946411133, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8782956600189209, "num_tokens": 663243166.0, "step": 17384 }, { "epoch": 2.2115506932960183, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.611189842224121, "learning_rate": 1e-06, "loss": 0.3326, "mean_token_accuracy": 0.8951841592788696, "num_tokens": 663282267.0, "step": 17385 }, { "epoch": 2.211677903574609, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.742161750793457, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8660292029380798, "num_tokens": 663318620.0, "step": 17386 }, { "epoch": 2.2118051138531993, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.673407554626465, "learning_rate": 1e-06, "loss": 0.3409, "mean_token_accuracy": 0.8914878368377686, "num_tokens": 663353256.0, "step": 17387 }, { "epoch": 2.21193232413179, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.649188041687012, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8768866062164307, "num_tokens": 663386691.0, "step": 17388 }, { "epoch": 2.2120595344103804, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.652261734008789, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8698351383209229, "num_tokens": 663416550.0, "step": 17389 }, { "epoch": 2.212186744688971, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.577284812927246, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8759838342666626, "num_tokens": 663464912.0, "step": 17390 }, { "epoch": 2.2123139549675614, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.662633895874023, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8652079105377197, "num_tokens": 663503786.0, "step": 17391 }, { "epoch": 2.212441165246152, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.6688871383667, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8819912672042847, "num_tokens": 663540579.0, "step": 17392 }, { "epoch": 2.2125683755247425, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.680087089538574, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8749645948410034, "num_tokens": 663577284.0, "step": 17393 }, { "epoch": 2.212695585803333, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.632914543151855, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8703467845916748, "num_tokens": 663613118.0, "step": 17394 }, { "epoch": 2.2128227960819236, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.700794219970703, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8777880072593689, "num_tokens": 663651548.0, "step": 17395 }, { "epoch": 2.212950006360514, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.661888122558594, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8568455576896667, "num_tokens": 663694381.0, "step": 17396 }, { "epoch": 2.2130772166391046, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.62879753112793, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.874436616897583, "num_tokens": 663734686.0, "step": 17397 }, { "epoch": 2.213204426917695, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.752397537231445, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8654689788818359, "num_tokens": 663768764.0, "step": 17398 }, { "epoch": 2.2133316371962852, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.664164543151855, "learning_rate": 1e-06, "loss": 0.3387, "mean_token_accuracy": 0.8934991359710693, "num_tokens": 663807650.0, "step": 17399 }, { "epoch": 2.213458847474876, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.732647895812988, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8711473345756531, "num_tokens": 663849308.0, "step": 17400 }, { "epoch": 2.2135860577534663, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.662503242492676, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8597807884216309, "num_tokens": 663888917.0, "step": 17401 }, { "epoch": 2.213713268032057, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.790457725524902, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8749791383743286, "num_tokens": 663921048.0, "step": 17402 }, { "epoch": 2.2138404783106473, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.628913879394531, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.879032552242279, "num_tokens": 663963433.0, "step": 17403 }, { "epoch": 2.213967688589238, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.721990585327148, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8565031290054321, "num_tokens": 664003294.0, "step": 17404 }, { "epoch": 2.2140948988678284, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.637907028198242, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.861945629119873, "num_tokens": 664043161.0, "step": 17405 }, { "epoch": 2.214222109146419, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.71998405456543, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8845169544219971, "num_tokens": 664082457.0, "step": 17406 }, { "epoch": 2.2143493194250095, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.68226432800293, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8836395740509033, "num_tokens": 664125999.0, "step": 17407 }, { "epoch": 2.2144765297036, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.680489540100098, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8793058395385742, "num_tokens": 664167256.0, "step": 17408 }, { "epoch": 2.2146037399821905, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.646798133850098, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.869051992893219, "num_tokens": 664201887.0, "step": 17409 }, { "epoch": 2.214730950260781, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.765660285949707, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8786613345146179, "num_tokens": 664238986.0, "step": 17410 }, { "epoch": 2.2148581605393716, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.618879318237305, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8697371482849121, "num_tokens": 664283436.0, "step": 17411 }, { "epoch": 2.214985370817962, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.765911102294922, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8675249218940735, "num_tokens": 664325196.0, "step": 17412 }, { "epoch": 2.2151125810965526, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.666854858398438, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8642323613166809, "num_tokens": 664365722.0, "step": 17413 }, { "epoch": 2.215239791375143, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.681896209716797, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8651818037033081, "num_tokens": 664411570.0, "step": 17414 }, { "epoch": 2.2153670016537337, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.690114974975586, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8765417337417603, "num_tokens": 664453545.0, "step": 17415 }, { "epoch": 2.215494211932324, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.733011245727539, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8744519948959351, "num_tokens": 664489882.0, "step": 17416 }, { "epoch": 2.2156214222109147, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.634449005126953, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.871724545955658, "num_tokens": 664528720.0, "step": 17417 }, { "epoch": 2.2157486324895053, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.603442192077637, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8646956086158752, "num_tokens": 664569645.0, "step": 17418 }, { "epoch": 2.215875842768096, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.709168434143066, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8685444593429565, "num_tokens": 664612033.0, "step": 17419 }, { "epoch": 2.2160030530466863, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.673418045043945, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8816531896591187, "num_tokens": 664651408.0, "step": 17420 }, { "epoch": 2.216130263325277, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.594773292541504, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8585612773895264, "num_tokens": 664695764.0, "step": 17421 }, { "epoch": 2.2162574736038674, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.661012649536133, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8770167827606201, "num_tokens": 664735528.0, "step": 17422 }, { "epoch": 2.216384683882458, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.644509315490723, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8657143115997314, "num_tokens": 664775612.0, "step": 17423 }, { "epoch": 2.216511894161048, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.651529312133789, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8696957230567932, "num_tokens": 664812485.0, "step": 17424 }, { "epoch": 2.2166391044396385, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.658792495727539, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8593703508377075, "num_tokens": 664853166.0, "step": 17425 }, { "epoch": 2.216766314718229, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.653048515319824, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8620761632919312, "num_tokens": 664888778.0, "step": 17426 }, { "epoch": 2.2168935249968196, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.65204906463623, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8716199398040771, "num_tokens": 664933783.0, "step": 17427 }, { "epoch": 2.21702073527541, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.633501052856445, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8563423156738281, "num_tokens": 664965228.0, "step": 17428 }, { "epoch": 2.2171479455540006, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.692586898803711, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8690555095672607, "num_tokens": 664996691.0, "step": 17429 }, { "epoch": 2.217275155832591, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.681968688964844, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8831120133399963, "num_tokens": 665032006.0, "step": 17430 }, { "epoch": 2.2174023661111817, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.616225242614746, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.862777590751648, "num_tokens": 665071819.0, "step": 17431 }, { "epoch": 2.217529576389772, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.630875587463379, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8620035648345947, "num_tokens": 665118883.0, "step": 17432 }, { "epoch": 2.2176567866683627, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.750838279724121, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8547841310501099, "num_tokens": 665154926.0, "step": 17433 }, { "epoch": 2.2177839969469533, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.650191307067871, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8815768957138062, "num_tokens": 665189910.0, "step": 17434 }, { "epoch": 2.217911207225544, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.60433292388916, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8794515132904053, "num_tokens": 665235849.0, "step": 17435 }, { "epoch": 2.2180384175041343, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.614882469177246, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8641371130943298, "num_tokens": 665274231.0, "step": 17436 }, { "epoch": 2.218165627782725, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.697318077087402, "learning_rate": 1e-06, "loss": 0.3433, "mean_token_accuracy": 0.8912019729614258, "num_tokens": 665309057.0, "step": 17437 }, { "epoch": 2.2182928380613154, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.621108055114746, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8702375888824463, "num_tokens": 665337031.0, "step": 17438 }, { "epoch": 2.218420048339906, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.660542488098145, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8723446130752563, "num_tokens": 665376557.0, "step": 17439 }, { "epoch": 2.2185472586184964, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.648884773254395, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8708127737045288, "num_tokens": 665419928.0, "step": 17440 }, { "epoch": 2.218674468897087, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.671523094177246, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8790408968925476, "num_tokens": 665458827.0, "step": 17441 }, { "epoch": 2.2188016791756775, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.635102272033691, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8753653764724731, "num_tokens": 665498623.0, "step": 17442 }, { "epoch": 2.218928889454268, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.716553688049316, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8683249950408936, "num_tokens": 665536103.0, "step": 17443 }, { "epoch": 2.2190560997328586, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.629020690917969, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8713502883911133, "num_tokens": 665572089.0, "step": 17444 }, { "epoch": 2.219183310011449, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.705455780029297, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8642425537109375, "num_tokens": 665608088.0, "step": 17445 }, { "epoch": 2.2193105202900396, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.724188804626465, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.881981611251831, "num_tokens": 665641883.0, "step": 17446 }, { "epoch": 2.21943773056863, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.685218811035156, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8771006464958191, "num_tokens": 665682211.0, "step": 17447 }, { "epoch": 2.2195649408472207, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.702993392944336, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8776721954345703, "num_tokens": 665726182.0, "step": 17448 }, { "epoch": 2.2196921511258108, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.6817045211792, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8764625191688538, "num_tokens": 665760404.0, "step": 17449 }, { "epoch": 2.2198193614044013, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.71407413482666, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8726146221160889, "num_tokens": 665798235.0, "step": 17450 }, { "epoch": 2.219946571682992, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.711019515991211, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8724939823150635, "num_tokens": 665840118.0, "step": 17451 }, { "epoch": 2.2200737819615823, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.641484260559082, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8722307682037354, "num_tokens": 665878498.0, "step": 17452 }, { "epoch": 2.220200992240173, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.618659973144531, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8687272071838379, "num_tokens": 665924179.0, "step": 17453 }, { "epoch": 2.2203282025187634, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.75963306427002, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8862942457199097, "num_tokens": 665963160.0, "step": 17454 }, { "epoch": 2.220455412797354, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.636977195739746, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.882076621055603, "num_tokens": 665998854.0, "step": 17455 }, { "epoch": 2.2205826230759445, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.71705150604248, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8813891410827637, "num_tokens": 666036420.0, "step": 17456 }, { "epoch": 2.220709833354535, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.74566650390625, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8537545204162598, "num_tokens": 666076692.0, "step": 17457 }, { "epoch": 2.2208370436331255, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.625249862670898, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.859269917011261, "num_tokens": 666118009.0, "step": 17458 }, { "epoch": 2.220964253911716, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.791685104370117, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8630142211914062, "num_tokens": 666158132.0, "step": 17459 }, { "epoch": 2.2210914641903066, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.730084419250488, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8676722049713135, "num_tokens": 666187195.0, "step": 17460 }, { "epoch": 2.221218674468897, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.725666046142578, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8781312704086304, "num_tokens": 666227790.0, "step": 17461 }, { "epoch": 2.2213458847474876, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.6520357131958, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8662700653076172, "num_tokens": 666268517.0, "step": 17462 }, { "epoch": 2.221473095026078, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.775670051574707, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8894309997558594, "num_tokens": 666313727.0, "step": 17463 }, { "epoch": 2.2216003053046687, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.676328659057617, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8761637210845947, "num_tokens": 666352713.0, "step": 17464 }, { "epoch": 2.221727515583259, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.716519355773926, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8806688189506531, "num_tokens": 666395065.0, "step": 17465 }, { "epoch": 2.2218547258618497, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.712886810302734, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8749446272850037, "num_tokens": 666428019.0, "step": 17466 }, { "epoch": 2.2219819361404403, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.743559837341309, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8721281290054321, "num_tokens": 666466830.0, "step": 17467 }, { "epoch": 2.222109146419031, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.723859786987305, "learning_rate": 1e-06, "loss": 0.3436, "mean_token_accuracy": 0.8903322219848633, "num_tokens": 666501028.0, "step": 17468 }, { "epoch": 2.2222363566976213, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.675694465637207, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8858683109283447, "num_tokens": 666538554.0, "step": 17469 }, { "epoch": 2.222363566976212, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.735444068908691, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.872897744178772, "num_tokens": 666577019.0, "step": 17470 }, { "epoch": 2.2224907772548024, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.68343734741211, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.887094259262085, "num_tokens": 666616597.0, "step": 17471 }, { "epoch": 2.222617987533393, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.679433822631836, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8774324655532837, "num_tokens": 666651523.0, "step": 17472 }, { "epoch": 2.2227451978119834, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.775066375732422, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8640490770339966, "num_tokens": 666689270.0, "step": 17473 }, { "epoch": 2.2228724080905735, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.693910598754883, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8694692850112915, "num_tokens": 666726836.0, "step": 17474 }, { "epoch": 2.222999618369164, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.723589897155762, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8700925707817078, "num_tokens": 666761207.0, "step": 17475 }, { "epoch": 2.2231268286477546, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.687464714050293, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8819835782051086, "num_tokens": 666800293.0, "step": 17476 }, { "epoch": 2.223254038926345, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.716833114624023, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8813000917434692, "num_tokens": 666843046.0, "step": 17477 }, { "epoch": 2.2233812492049356, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.590795516967773, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.88653963804245, "num_tokens": 666885840.0, "step": 17478 }, { "epoch": 2.223508459483526, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.717293739318848, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8740919828414917, "num_tokens": 666925654.0, "step": 17479 }, { "epoch": 2.2236356697621167, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.714524269104004, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8752442598342896, "num_tokens": 666968098.0, "step": 17480 }, { "epoch": 2.223762880040707, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.643949508666992, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8837378025054932, "num_tokens": 667010349.0, "step": 17481 }, { "epoch": 2.2238900903192977, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.69554328918457, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8652749061584473, "num_tokens": 667047086.0, "step": 17482 }, { "epoch": 2.2240173005978883, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.750405311584473, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8574349880218506, "num_tokens": 667083955.0, "step": 17483 }, { "epoch": 2.224144510876479, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.696441650390625, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8802651762962341, "num_tokens": 667119751.0, "step": 17484 }, { "epoch": 2.2242717211550693, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.696295738220215, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8693519234657288, "num_tokens": 667150980.0, "step": 17485 }, { "epoch": 2.22439893143366, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.744421005249023, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8707391619682312, "num_tokens": 667185198.0, "step": 17486 }, { "epoch": 2.2245261417122504, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.658860206604004, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8763922452926636, "num_tokens": 667225286.0, "step": 17487 }, { "epoch": 2.224653351990841, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.630462646484375, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8675463199615479, "num_tokens": 667266294.0, "step": 17488 }, { "epoch": 2.2247805622694314, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.682722091674805, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8709403872489929, "num_tokens": 667302430.0, "step": 17489 }, { "epoch": 2.224907772548022, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.652484893798828, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8678746223449707, "num_tokens": 667342881.0, "step": 17490 }, { "epoch": 2.2250349828266125, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.712620735168457, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8756489157676697, "num_tokens": 667376122.0, "step": 17491 }, { "epoch": 2.225162193105203, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.686724662780762, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8672274351119995, "num_tokens": 667415120.0, "step": 17492 }, { "epoch": 2.2252894033837936, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.61191177368164, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.87321537733078, "num_tokens": 667451014.0, "step": 17493 }, { "epoch": 2.225416613662384, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.706055641174316, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8639284372329712, "num_tokens": 667485600.0, "step": 17494 }, { "epoch": 2.2255438239409746, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.656805038452148, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8688006401062012, "num_tokens": 667527561.0, "step": 17495 }, { "epoch": 2.225671034219565, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.676997184753418, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8806102275848389, "num_tokens": 667569677.0, "step": 17496 }, { "epoch": 2.225798244498155, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.70030403137207, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8726876378059387, "num_tokens": 667605954.0, "step": 17497 }, { "epoch": 2.225925454776746, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.69102954864502, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8757503032684326, "num_tokens": 667645082.0, "step": 17498 }, { "epoch": 2.2260526650553363, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.627138137817383, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8699026107788086, "num_tokens": 667681390.0, "step": 17499 }, { "epoch": 2.226179875333927, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.743172645568848, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.86278235912323, "num_tokens": 667722772.0, "step": 17500 }, { "epoch": 2.2263070856125173, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.622651100158691, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8479368686676025, "num_tokens": 667758647.0, "step": 17501 }, { "epoch": 2.226434295891108, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.756300926208496, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8738825917243958, "num_tokens": 667790625.0, "step": 17502 }, { "epoch": 2.2265615061696984, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.77106761932373, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8634912371635437, "num_tokens": 667823724.0, "step": 17503 }, { "epoch": 2.226688716448289, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.639636993408203, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8861647844314575, "num_tokens": 667862548.0, "step": 17504 }, { "epoch": 2.2268159267268794, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.651981353759766, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8661320209503174, "num_tokens": 667905994.0, "step": 17505 }, { "epoch": 2.22694313700547, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.729412078857422, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8668431043624878, "num_tokens": 667945087.0, "step": 17506 }, { "epoch": 2.2270703472840605, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.618715286254883, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8823766708374023, "num_tokens": 667984736.0, "step": 17507 }, { "epoch": 2.227197557562651, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.767416954040527, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8674596548080444, "num_tokens": 668024020.0, "step": 17508 }, { "epoch": 2.2273247678412416, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.708606719970703, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8682414889335632, "num_tokens": 668059567.0, "step": 17509 }, { "epoch": 2.227451978119832, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.671512603759766, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8792625665664673, "num_tokens": 668099901.0, "step": 17510 }, { "epoch": 2.2275791883984226, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.767925262451172, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8871550559997559, "num_tokens": 668135974.0, "step": 17511 }, { "epoch": 2.227706398677013, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.686817169189453, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8752070069313049, "num_tokens": 668179200.0, "step": 17512 }, { "epoch": 2.2278336089556037, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.737070083618164, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8516098260879517, "num_tokens": 668219683.0, "step": 17513 }, { "epoch": 2.227960819234194, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.72960376739502, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8753618001937866, "num_tokens": 668253899.0, "step": 17514 }, { "epoch": 2.2280880295127847, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.786825180053711, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8670158386230469, "num_tokens": 668294140.0, "step": 17515 }, { "epoch": 2.2282152397913753, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.688220024108887, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8799751400947571, "num_tokens": 668328442.0, "step": 17516 }, { "epoch": 2.228342450069966, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.678088188171387, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8764563798904419, "num_tokens": 668367903.0, "step": 17517 }, { "epoch": 2.2284696603485563, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.785706520080566, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8854119777679443, "num_tokens": 668409889.0, "step": 17518 }, { "epoch": 2.228596870627147, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.731727600097656, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8767789006233215, "num_tokens": 668451061.0, "step": 17519 }, { "epoch": 2.2287240809057374, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.687610626220703, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8514593839645386, "num_tokens": 668493038.0, "step": 17520 }, { "epoch": 2.228851291184328, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.742706298828125, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8807080388069153, "num_tokens": 668528087.0, "step": 17521 }, { "epoch": 2.228978501462918, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.722496032714844, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8745540380477905, "num_tokens": 668562325.0, "step": 17522 }, { "epoch": 2.2291057117415085, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.707125663757324, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8763068914413452, "num_tokens": 668601336.0, "step": 17523 }, { "epoch": 2.229232922020099, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.705606460571289, "learning_rate": 1e-06, "loss": 0.3345, "mean_token_accuracy": 0.8915587663650513, "num_tokens": 668637113.0, "step": 17524 }, { "epoch": 2.2293601322986896, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.706578254699707, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8681737184524536, "num_tokens": 668675645.0, "step": 17525 }, { "epoch": 2.22948734257728, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.771336555480957, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8746955394744873, "num_tokens": 668706480.0, "step": 17526 }, { "epoch": 2.2296145528558706, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.777928352355957, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8676093816757202, "num_tokens": 668743956.0, "step": 17527 }, { "epoch": 2.229741763134461, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.719367027282715, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8822650909423828, "num_tokens": 668779134.0, "step": 17528 }, { "epoch": 2.2298689734130517, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.735260963439941, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8755697011947632, "num_tokens": 668814978.0, "step": 17529 }, { "epoch": 2.229996183691642, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.685810089111328, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8580240607261658, "num_tokens": 668854194.0, "step": 17530 }, { "epoch": 2.2301233939702327, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.838722229003906, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8654429912567139, "num_tokens": 668893095.0, "step": 17531 }, { "epoch": 2.2302506042488233, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.714346885681152, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8800102472305298, "num_tokens": 668930626.0, "step": 17532 }, { "epoch": 2.230377814527414, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.697013854980469, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8806637525558472, "num_tokens": 668960941.0, "step": 17533 }, { "epoch": 2.2305050248060043, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.760954856872559, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.876136064529419, "num_tokens": 668997771.0, "step": 17534 }, { "epoch": 2.230632235084595, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.778061866760254, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8865670561790466, "num_tokens": 669038714.0, "step": 17535 }, { "epoch": 2.2307594453631854, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.746582984924316, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8827481865882874, "num_tokens": 669072336.0, "step": 17536 }, { "epoch": 2.230886655641776, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.666976928710938, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8834977746009827, "num_tokens": 669110303.0, "step": 17537 }, { "epoch": 2.2310138659203664, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.756775856018066, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.863368034362793, "num_tokens": 669146874.0, "step": 17538 }, { "epoch": 2.231141076198957, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.750476837158203, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8853330612182617, "num_tokens": 669181682.0, "step": 17539 }, { "epoch": 2.2312682864775475, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.766258239746094, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8731769323348999, "num_tokens": 669227201.0, "step": 17540 }, { "epoch": 2.231395496756138, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.702853202819824, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8734901547431946, "num_tokens": 669265915.0, "step": 17541 }, { "epoch": 2.2315227070347285, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.77104663848877, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8778828978538513, "num_tokens": 669304130.0, "step": 17542 }, { "epoch": 2.231649917313319, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.76488971710205, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8717811703681946, "num_tokens": 669344225.0, "step": 17543 }, { "epoch": 2.2317771275919096, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.799860954284668, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.856175422668457, "num_tokens": 669385126.0, "step": 17544 }, { "epoch": 2.2319043378705, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.683695793151855, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8618863224983215, "num_tokens": 669425774.0, "step": 17545 }, { "epoch": 2.2320315481490907, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.792771339416504, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8641848564147949, "num_tokens": 669459789.0, "step": 17546 }, { "epoch": 2.2321587584276807, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.719752311706543, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8881158828735352, "num_tokens": 669498709.0, "step": 17547 }, { "epoch": 2.2322859687062713, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.75684928894043, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8779650330543518, "num_tokens": 669536021.0, "step": 17548 }, { "epoch": 2.232413178984862, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.7407865524292, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8614440560340881, "num_tokens": 669574316.0, "step": 17549 }, { "epoch": 2.2325403892634523, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.773846626281738, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8639122843742371, "num_tokens": 669611826.0, "step": 17550 }, { "epoch": 2.232667599542043, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.784954071044922, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8704957365989685, "num_tokens": 669655277.0, "step": 17551 }, { "epoch": 2.2327948098206334, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.74984359741211, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8665857911109924, "num_tokens": 669693481.0, "step": 17552 }, { "epoch": 2.232922020099224, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.781310081481934, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8731407523155212, "num_tokens": 669724104.0, "step": 17553 }, { "epoch": 2.2330492303778144, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.66818904876709, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8559974431991577, "num_tokens": 669768725.0, "step": 17554 }, { "epoch": 2.233176440656405, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.741349220275879, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8815982341766357, "num_tokens": 669809508.0, "step": 17555 }, { "epoch": 2.2333036509349955, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.643086433410645, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8741228580474854, "num_tokens": 669848047.0, "step": 17556 }, { "epoch": 2.233430861213586, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.740480422973633, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8890585899353027, "num_tokens": 669881274.0, "step": 17557 }, { "epoch": 2.2335580714921766, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.694093704223633, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8671640753746033, "num_tokens": 669918517.0, "step": 17558 }, { "epoch": 2.233685281770767, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.708014488220215, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8654471635818481, "num_tokens": 669960990.0, "step": 17559 }, { "epoch": 2.2338124920493576, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.662186622619629, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8782764673233032, "num_tokens": 669995542.0, "step": 17560 }, { "epoch": 2.233939702327948, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.72797679901123, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8746311664581299, "num_tokens": 670036798.0, "step": 17561 }, { "epoch": 2.2340669126065387, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.702560424804688, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.87158203125, "num_tokens": 670071763.0, "step": 17562 }, { "epoch": 2.234194122885129, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.713798522949219, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8727157711982727, "num_tokens": 670100864.0, "step": 17563 }, { "epoch": 2.2343213331637197, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.640948295593262, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8616366386413574, "num_tokens": 670144114.0, "step": 17564 }, { "epoch": 2.2344485434423103, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.703824996948242, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8772042989730835, "num_tokens": 670182719.0, "step": 17565 }, { "epoch": 2.234575753720901, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.723054885864258, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8768208026885986, "num_tokens": 670218059.0, "step": 17566 }, { "epoch": 2.2347029639994913, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.680517196655273, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.869392454624176, "num_tokens": 670259589.0, "step": 17567 }, { "epoch": 2.234830174278082, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.655106544494629, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.864956259727478, "num_tokens": 670301755.0, "step": 17568 }, { "epoch": 2.2349573845566724, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.755765914916992, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8775347471237183, "num_tokens": 670339205.0, "step": 17569 }, { "epoch": 2.235084594835263, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.723533630371094, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8756825923919678, "num_tokens": 670382961.0, "step": 17570 }, { "epoch": 2.2352118051138534, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.707554817199707, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8628983497619629, "num_tokens": 670420747.0, "step": 17571 }, { "epoch": 2.2353390153924435, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.692233085632324, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8899343013763428, "num_tokens": 670457906.0, "step": 17572 }, { "epoch": 2.235466225671034, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.733342170715332, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8615710735321045, "num_tokens": 670493187.0, "step": 17573 }, { "epoch": 2.2355934359496246, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.697724342346191, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8865187764167786, "num_tokens": 670529750.0, "step": 17574 }, { "epoch": 2.235720646228215, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.71966552734375, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8709499835968018, "num_tokens": 670570147.0, "step": 17575 }, { "epoch": 2.2358478565068056, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.679508209228516, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8744409084320068, "num_tokens": 670604113.0, "step": 17576 }, { "epoch": 2.235975066785396, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.733735084533691, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8724485635757446, "num_tokens": 670645451.0, "step": 17577 }, { "epoch": 2.2361022770639867, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.706210136413574, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8780124187469482, "num_tokens": 670684055.0, "step": 17578 }, { "epoch": 2.236229487342577, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.741695404052734, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8644395470619202, "num_tokens": 670724397.0, "step": 17579 }, { "epoch": 2.2363566976211677, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.734484672546387, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8835543394088745, "num_tokens": 670761282.0, "step": 17580 }, { "epoch": 2.2364839078997583, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.764090538024902, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8773486614227295, "num_tokens": 670800007.0, "step": 17581 }, { "epoch": 2.236611118178349, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.746773719787598, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8702949285507202, "num_tokens": 670840040.0, "step": 17582 }, { "epoch": 2.2367383284569393, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.712422370910645, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.865890383720398, "num_tokens": 670881193.0, "step": 17583 }, { "epoch": 2.23686553873553, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.72036075592041, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8766687512397766, "num_tokens": 670927275.0, "step": 17584 }, { "epoch": 2.2369927490141204, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.710335731506348, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.877045750617981, "num_tokens": 670967327.0, "step": 17585 }, { "epoch": 2.237119959292711, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.674689292907715, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8770240545272827, "num_tokens": 671009250.0, "step": 17586 }, { "epoch": 2.2372471695713014, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.745940208435059, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.877010703086853, "num_tokens": 671045068.0, "step": 17587 }, { "epoch": 2.237374379849892, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.681894302368164, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8608052730560303, "num_tokens": 671084786.0, "step": 17588 }, { "epoch": 2.2375015901284825, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.776527404785156, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8665581941604614, "num_tokens": 671123235.0, "step": 17589 }, { "epoch": 2.237628800407073, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.776658058166504, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8892093896865845, "num_tokens": 671161702.0, "step": 17590 }, { "epoch": 2.2377560106856635, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.742395401000977, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8845286965370178, "num_tokens": 671195487.0, "step": 17591 }, { "epoch": 2.237883220964254, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.721635818481445, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8709178566932678, "num_tokens": 671232288.0, "step": 17592 }, { "epoch": 2.2380104312428446, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.695526123046875, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8564469814300537, "num_tokens": 671275154.0, "step": 17593 }, { "epoch": 2.238137641521435, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.834610939025879, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8628633618354797, "num_tokens": 671312266.0, "step": 17594 }, { "epoch": 2.238264851800025, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.737610816955566, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.872681736946106, "num_tokens": 671349690.0, "step": 17595 }, { "epoch": 2.238392062078616, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.750677108764648, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8771296739578247, "num_tokens": 671387898.0, "step": 17596 }, { "epoch": 2.2385192723572063, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.753029823303223, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8523521423339844, "num_tokens": 671428054.0, "step": 17597 }, { "epoch": 2.238646482635797, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.693370819091797, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8792707920074463, "num_tokens": 671467096.0, "step": 17598 }, { "epoch": 2.2387736929143873, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.693288803100586, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8720398545265198, "num_tokens": 671506752.0, "step": 17599 }, { "epoch": 2.238900903192978, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.768973350524902, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8657784461975098, "num_tokens": 671542403.0, "step": 17600 }, { "epoch": 2.2390281134715684, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.743168830871582, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8846310377120972, "num_tokens": 671578115.0, "step": 17601 }, { "epoch": 2.239155323750159, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.708847045898438, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8834421038627625, "num_tokens": 671619854.0, "step": 17602 }, { "epoch": 2.2392825340287494, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.674768447875977, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8868108987808228, "num_tokens": 671661617.0, "step": 17603 }, { "epoch": 2.23940974430734, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.732290267944336, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8624497652053833, "num_tokens": 671702231.0, "step": 17604 }, { "epoch": 2.2395369545859305, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.684473991394043, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.873709499835968, "num_tokens": 671741483.0, "step": 17605 }, { "epoch": 2.239664164864521, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.715929985046387, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8693551421165466, "num_tokens": 671780168.0, "step": 17606 }, { "epoch": 2.2397913751431116, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.701908111572266, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8504112958908081, "num_tokens": 671818980.0, "step": 17607 }, { "epoch": 2.239918585421702, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.725903511047363, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8731144666671753, "num_tokens": 671850010.0, "step": 17608 }, { "epoch": 2.2400457957002926, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.77298355102539, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8656362295150757, "num_tokens": 671888506.0, "step": 17609 }, { "epoch": 2.240173005978883, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.697380065917969, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8818019032478333, "num_tokens": 671930965.0, "step": 17610 }, { "epoch": 2.2403002162574737, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.799644470214844, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8815330266952515, "num_tokens": 671967472.0, "step": 17611 }, { "epoch": 2.240427426536064, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.760499954223633, "learning_rate": 1e-06, "loss": 0.3268, "mean_token_accuracy": 0.895286500453949, "num_tokens": 672005986.0, "step": 17612 }, { "epoch": 2.2405546368146547, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.813075065612793, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8734246492385864, "num_tokens": 672044729.0, "step": 17613 }, { "epoch": 2.2406818470932452, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.789871215820312, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8738406896591187, "num_tokens": 672080141.0, "step": 17614 }, { "epoch": 2.2408090573718358, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.813837051391602, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8808344006538391, "num_tokens": 672119277.0, "step": 17615 }, { "epoch": 2.2409362676504263, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.764776229858398, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8712533712387085, "num_tokens": 672154205.0, "step": 17616 }, { "epoch": 2.241063477929017, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.842351913452148, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8728054165840149, "num_tokens": 672193667.0, "step": 17617 }, { "epoch": 2.2411906882076074, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.784558296203613, "learning_rate": 1e-06, "loss": 0.4942, "mean_token_accuracy": 0.8441617488861084, "num_tokens": 672232360.0, "step": 17618 }, { "epoch": 2.241317898486198, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.772815704345703, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8607714176177979, "num_tokens": 672264127.0, "step": 17619 }, { "epoch": 2.241445108764788, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.79164981842041, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8808836936950684, "num_tokens": 672300719.0, "step": 17620 }, { "epoch": 2.2415723190433785, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.699047088623047, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8836563229560852, "num_tokens": 672342796.0, "step": 17621 }, { "epoch": 2.241699529321969, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.840920448303223, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8747302293777466, "num_tokens": 672382597.0, "step": 17622 }, { "epoch": 2.2418267396005596, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.73738956451416, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8614264726638794, "num_tokens": 672427350.0, "step": 17623 }, { "epoch": 2.24195394987915, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.759462356567383, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8669155240058899, "num_tokens": 672468997.0, "step": 17624 }, { "epoch": 2.2420811601577406, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.741686820983887, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8792311549186707, "num_tokens": 672505061.0, "step": 17625 }, { "epoch": 2.242208370436331, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.746018409729004, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8716284036636353, "num_tokens": 672544241.0, "step": 17626 }, { "epoch": 2.2423355807149217, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.845952987670898, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8714818954467773, "num_tokens": 672586334.0, "step": 17627 }, { "epoch": 2.242462790993512, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.741996765136719, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8694152235984802, "num_tokens": 672628627.0, "step": 17628 }, { "epoch": 2.2425900012721027, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.69733715057373, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8639053106307983, "num_tokens": 672663046.0, "step": 17629 }, { "epoch": 2.2427172115506933, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.671295166015625, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8792190551757812, "num_tokens": 672703394.0, "step": 17630 }, { "epoch": 2.242844421829284, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.712339401245117, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8642966747283936, "num_tokens": 672743262.0, "step": 17631 }, { "epoch": 2.2429716321078743, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.767671585083008, "learning_rate": 1e-06, "loss": 0.4963, "mean_token_accuracy": 0.8398562669754028, "num_tokens": 672782662.0, "step": 17632 }, { "epoch": 2.243098842386465, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.774709701538086, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8701914548873901, "num_tokens": 672818400.0, "step": 17633 }, { "epoch": 2.2432260526650554, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.785234451293945, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.885587751865387, "num_tokens": 672858843.0, "step": 17634 }, { "epoch": 2.243353262943646, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.65774917602539, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8743605613708496, "num_tokens": 672900397.0, "step": 17635 }, { "epoch": 2.2434804732222364, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.796256065368652, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.871059775352478, "num_tokens": 672936025.0, "step": 17636 }, { "epoch": 2.243607683500827, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.764843940734863, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8545936346054077, "num_tokens": 672979180.0, "step": 17637 }, { "epoch": 2.2437348937794175, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.67478084564209, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8826891183853149, "num_tokens": 673018053.0, "step": 17638 }, { "epoch": 2.243862104058008, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.761035919189453, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8540223836898804, "num_tokens": 673057596.0, "step": 17639 }, { "epoch": 2.2439893143365985, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.731738090515137, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8847487568855286, "num_tokens": 673097740.0, "step": 17640 }, { "epoch": 2.244116524615189, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.696173667907715, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8781436681747437, "num_tokens": 673141821.0, "step": 17641 }, { "epoch": 2.2442437348937796, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.78760814666748, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8805409669876099, "num_tokens": 673184279.0, "step": 17642 }, { "epoch": 2.24437094517237, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.683568954467773, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8825339674949646, "num_tokens": 673224328.0, "step": 17643 }, { "epoch": 2.2444981554509607, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.735328674316406, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8597252368927002, "num_tokens": 673263328.0, "step": 17644 }, { "epoch": 2.2446253657295507, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.782767295837402, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8771424293518066, "num_tokens": 673301663.0, "step": 17645 }, { "epoch": 2.2447525760081413, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.721898078918457, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8705912828445435, "num_tokens": 673340606.0, "step": 17646 }, { "epoch": 2.244879786286732, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.75381851196289, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8706095218658447, "num_tokens": 673382201.0, "step": 17647 }, { "epoch": 2.2450069965653223, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.68362045288086, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.882239580154419, "num_tokens": 673423330.0, "step": 17648 }, { "epoch": 2.245134206843913, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.85107135772705, "learning_rate": 1e-06, "loss": 0.3402, "mean_token_accuracy": 0.8888540863990784, "num_tokens": 673459690.0, "step": 17649 }, { "epoch": 2.2452614171225034, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.726704597473145, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8682814836502075, "num_tokens": 673494827.0, "step": 17650 }, { "epoch": 2.245388627401094, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.743854522705078, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8609309196472168, "num_tokens": 673540568.0, "step": 17651 }, { "epoch": 2.2455158376796844, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.755189895629883, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.888534426689148, "num_tokens": 673574022.0, "step": 17652 }, { "epoch": 2.245643047958275, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.791242599487305, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8692110776901245, "num_tokens": 673612493.0, "step": 17653 }, { "epoch": 2.2457702582368655, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.73305892944336, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8687682151794434, "num_tokens": 673653729.0, "step": 17654 }, { "epoch": 2.245897468515456, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.755155563354492, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8860431909561157, "num_tokens": 673695626.0, "step": 17655 }, { "epoch": 2.2460246787940465, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.748376846313477, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8795083165168762, "num_tokens": 673732188.0, "step": 17656 }, { "epoch": 2.246151889072637, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.776650428771973, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8644459247589111, "num_tokens": 673770259.0, "step": 17657 }, { "epoch": 2.2462790993512276, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.684552192687988, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8689948320388794, "num_tokens": 673805018.0, "step": 17658 }, { "epoch": 2.246406309629818, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.727621078491211, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8668681979179382, "num_tokens": 673846058.0, "step": 17659 }, { "epoch": 2.2465335199084087, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.6997709274292, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8594274520874023, "num_tokens": 673886437.0, "step": 17660 }, { "epoch": 2.246660730186999, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.694906234741211, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8757965564727783, "num_tokens": 673925701.0, "step": 17661 }, { "epoch": 2.2467879404655897, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.700113296508789, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8759592771530151, "num_tokens": 673969854.0, "step": 17662 }, { "epoch": 2.2469151507441802, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.725390434265137, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8751908540725708, "num_tokens": 674005048.0, "step": 17663 }, { "epoch": 2.2470423610227708, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.725728034973145, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.867866039276123, "num_tokens": 674046548.0, "step": 17664 }, { "epoch": 2.2471695713013613, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.710180282592773, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8669314384460449, "num_tokens": 674086905.0, "step": 17665 }, { "epoch": 2.247296781579952, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.756458282470703, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8827021718025208, "num_tokens": 674124589.0, "step": 17666 }, { "epoch": 2.2474239918585424, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.742669105529785, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8857553005218506, "num_tokens": 674165135.0, "step": 17667 }, { "epoch": 2.247551202137133, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.706168174743652, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.868503749370575, "num_tokens": 674207269.0, "step": 17668 }, { "epoch": 2.2476784124157234, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.757918357849121, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8889517188072205, "num_tokens": 674240806.0, "step": 17669 }, { "epoch": 2.2478056226943135, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.75529670715332, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8647102117538452, "num_tokens": 674274697.0, "step": 17670 }, { "epoch": 2.247932832972904, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.677539825439453, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8633776903152466, "num_tokens": 674311362.0, "step": 17671 }, { "epoch": 2.2480600432514946, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.752016067504883, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8797571063041687, "num_tokens": 674352505.0, "step": 17672 }, { "epoch": 2.248187253530085, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.791484832763672, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8782281875610352, "num_tokens": 674383938.0, "step": 17673 }, { "epoch": 2.2483144638086756, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.696195602416992, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.869811475276947, "num_tokens": 674421665.0, "step": 17674 }, { "epoch": 2.248441674087266, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.76437759399414, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8754833936691284, "num_tokens": 674466065.0, "step": 17675 }, { "epoch": 2.2485688843658567, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.667240142822266, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.874413788318634, "num_tokens": 674506868.0, "step": 17676 }, { "epoch": 2.248696094644447, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.69121265411377, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8690153360366821, "num_tokens": 674545756.0, "step": 17677 }, { "epoch": 2.2488233049230377, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.69881820678711, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8679037690162659, "num_tokens": 674590680.0, "step": 17678 }, { "epoch": 2.2489505152016283, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.733477592468262, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8781059980392456, "num_tokens": 674631624.0, "step": 17679 }, { "epoch": 2.249077725480219, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.664090156555176, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8676430583000183, "num_tokens": 674667365.0, "step": 17680 }, { "epoch": 2.2492049357588093, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.776338577270508, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8717595338821411, "num_tokens": 674706587.0, "step": 17681 }, { "epoch": 2.2493321460374, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.765239715576172, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8718748688697815, "num_tokens": 674744480.0, "step": 17682 }, { "epoch": 2.2494593563159904, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.735587120056152, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8726669549942017, "num_tokens": 674783104.0, "step": 17683 }, { "epoch": 2.249586566594581, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.7733736038208, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8656723499298096, "num_tokens": 674823488.0, "step": 17684 }, { "epoch": 2.2497137768731714, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.771307945251465, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8782928586006165, "num_tokens": 674857322.0, "step": 17685 }, { "epoch": 2.249840987151762, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.701313972473145, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8642833232879639, "num_tokens": 674901566.0, "step": 17686 }, { "epoch": 2.2499681974303525, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.77264404296875, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8783824443817139, "num_tokens": 674942803.0, "step": 17687 }, { "epoch": 2.250095407708943, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.708610534667969, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8648149967193604, "num_tokens": 674983160.0, "step": 17688 }, { "epoch": 2.2502226179875335, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.745410919189453, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8746081590652466, "num_tokens": 675021243.0, "step": 17689 }, { "epoch": 2.250349828266124, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.740885734558105, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8696268200874329, "num_tokens": 675059435.0, "step": 17690 }, { "epoch": 2.2504770385447146, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.780171394348145, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.875356137752533, "num_tokens": 675091689.0, "step": 17691 }, { "epoch": 2.250604248823305, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.76368522644043, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8674517869949341, "num_tokens": 675130368.0, "step": 17692 }, { "epoch": 2.250731459101895, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.727860450744629, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8760924339294434, "num_tokens": 675171745.0, "step": 17693 }, { "epoch": 2.250858669380486, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.831082344055176, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8718207478523254, "num_tokens": 675206052.0, "step": 17694 }, { "epoch": 2.2509858796590763, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.763787269592285, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8681713342666626, "num_tokens": 675243231.0, "step": 17695 }, { "epoch": 2.251113089937667, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.793152809143066, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8658362627029419, "num_tokens": 675273582.0, "step": 17696 }, { "epoch": 2.2512403002162573, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.695804595947266, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8693981170654297, "num_tokens": 675314247.0, "step": 17697 }, { "epoch": 2.251367510494848, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.740262031555176, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8739703893661499, "num_tokens": 675355675.0, "step": 17698 }, { "epoch": 2.2514947207734384, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.776082038879395, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8701558113098145, "num_tokens": 675394290.0, "step": 17699 }, { "epoch": 2.251621931052029, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.740466117858887, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8759495615959167, "num_tokens": 675434251.0, "step": 17700 }, { "epoch": 2.2517491413306194, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.741726875305176, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8790137767791748, "num_tokens": 675474731.0, "step": 17701 }, { "epoch": 2.25187635160921, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.732930183410645, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.867957592010498, "num_tokens": 675515180.0, "step": 17702 }, { "epoch": 2.2520035618878005, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.7728910446167, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8624703288078308, "num_tokens": 675553429.0, "step": 17703 }, { "epoch": 2.252130772166391, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.712045669555664, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8763094544410706, "num_tokens": 675592919.0, "step": 17704 }, { "epoch": 2.2522579824449815, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.782332420349121, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8803605437278748, "num_tokens": 675627528.0, "step": 17705 }, { "epoch": 2.252385192723572, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.654187202453613, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8809977769851685, "num_tokens": 675663721.0, "step": 17706 }, { "epoch": 2.2525124030021626, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.752412796020508, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8679307699203491, "num_tokens": 675705187.0, "step": 17707 }, { "epoch": 2.252639613280753, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.698028564453125, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8820075988769531, "num_tokens": 675752713.0, "step": 17708 }, { "epoch": 2.2527668235593437, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.803800582885742, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8743042945861816, "num_tokens": 675791009.0, "step": 17709 }, { "epoch": 2.252894033837934, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.72602653503418, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8761692047119141, "num_tokens": 675828890.0, "step": 17710 }, { "epoch": 2.2530212441165247, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.812065124511719, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8739485740661621, "num_tokens": 675862855.0, "step": 17711 }, { "epoch": 2.2531484543951152, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.76504135131836, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8712226152420044, "num_tokens": 675903032.0, "step": 17712 }, { "epoch": 2.2532756646737058, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.713664054870605, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.869293749332428, "num_tokens": 675943035.0, "step": 17713 }, { "epoch": 2.2534028749522963, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.740457534790039, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8754456043243408, "num_tokens": 675983073.0, "step": 17714 }, { "epoch": 2.253530085230887, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.696430206298828, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8783093690872192, "num_tokens": 676022341.0, "step": 17715 }, { "epoch": 2.2536572955094774, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.781661033630371, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8698033094406128, "num_tokens": 676060907.0, "step": 17716 }, { "epoch": 2.253784505788068, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.783583641052246, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8716630935668945, "num_tokens": 676100693.0, "step": 17717 }, { "epoch": 2.253911716066658, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.763616561889648, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8865605592727661, "num_tokens": 676137526.0, "step": 17718 }, { "epoch": 2.254038926345249, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.751379013061523, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8758020401000977, "num_tokens": 676177245.0, "step": 17719 }, { "epoch": 2.254166136623839, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.77772045135498, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8794001340866089, "num_tokens": 676214568.0, "step": 17720 }, { "epoch": 2.2542933469024296, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.764352798461914, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8742542862892151, "num_tokens": 676252552.0, "step": 17721 }, { "epoch": 2.25442055718102, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.798569679260254, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8557359576225281, "num_tokens": 676290480.0, "step": 17722 }, { "epoch": 2.2545477674596106, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.769952774047852, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8856431841850281, "num_tokens": 676324950.0, "step": 17723 }, { "epoch": 2.254674977738201, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.7689208984375, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.865382194519043, "num_tokens": 676363851.0, "step": 17724 }, { "epoch": 2.2548021880167917, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.794681549072266, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.882970929145813, "num_tokens": 676399920.0, "step": 17725 }, { "epoch": 2.254929398295382, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.718432426452637, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8704773187637329, "num_tokens": 676443243.0, "step": 17726 }, { "epoch": 2.2550566085739727, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.739684104919434, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8596400618553162, "num_tokens": 676480272.0, "step": 17727 }, { "epoch": 2.2551838188525632, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.66016674041748, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8684970140457153, "num_tokens": 676516630.0, "step": 17728 }, { "epoch": 2.2553110291311538, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.768576622009277, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8779355883598328, "num_tokens": 676557842.0, "step": 17729 }, { "epoch": 2.2554382394097443, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.702723503112793, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8835906386375427, "num_tokens": 676599413.0, "step": 17730 }, { "epoch": 2.255565449688335, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.79751968383789, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8655375838279724, "num_tokens": 676637791.0, "step": 17731 }, { "epoch": 2.2556926599669254, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.716792106628418, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.869714081287384, "num_tokens": 676677108.0, "step": 17732 }, { "epoch": 2.255819870245516, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.75558090209961, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8748447299003601, "num_tokens": 676720698.0, "step": 17733 }, { "epoch": 2.2559470805241064, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.741244316101074, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.862555742263794, "num_tokens": 676763719.0, "step": 17734 }, { "epoch": 2.256074290802697, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.713411331176758, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8737313747406006, "num_tokens": 676801083.0, "step": 17735 }, { "epoch": 2.2562015010812875, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.772047996520996, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8659145832061768, "num_tokens": 676840861.0, "step": 17736 }, { "epoch": 2.256328711359878, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.65231704711914, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.885830283164978, "num_tokens": 676876864.0, "step": 17737 }, { "epoch": 2.2564559216384685, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.748518943786621, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.8927050828933716, "num_tokens": 676920983.0, "step": 17738 }, { "epoch": 2.256583131917059, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.683691024780273, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.881271481513977, "num_tokens": 676961294.0, "step": 17739 }, { "epoch": 2.2567103421956496, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.755515098571777, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8653515577316284, "num_tokens": 676996318.0, "step": 17740 }, { "epoch": 2.2568375524742397, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.642945289611816, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8754357099533081, "num_tokens": 677034083.0, "step": 17741 }, { "epoch": 2.2569647627528306, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.748943328857422, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8750885128974915, "num_tokens": 677071206.0, "step": 17742 }, { "epoch": 2.2570919730314207, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.686537742614746, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8784154653549194, "num_tokens": 677103573.0, "step": 17743 }, { "epoch": 2.2572191833100113, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.751200675964355, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.882358729839325, "num_tokens": 677141406.0, "step": 17744 }, { "epoch": 2.257346393588602, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.734306335449219, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8627922534942627, "num_tokens": 677185968.0, "step": 17745 }, { "epoch": 2.2574736038671923, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.710688591003418, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8661381602287292, "num_tokens": 677229004.0, "step": 17746 }, { "epoch": 2.257600814145783, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.827000617980957, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8832371234893799, "num_tokens": 677266317.0, "step": 17747 }, { "epoch": 2.2577280244243734, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.742639541625977, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8832104206085205, "num_tokens": 677302344.0, "step": 17748 }, { "epoch": 2.257855234702964, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.767366409301758, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8639949560165405, "num_tokens": 677342103.0, "step": 17749 }, { "epoch": 2.2579824449815544, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.782722473144531, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.861479640007019, "num_tokens": 677378602.0, "step": 17750 }, { "epoch": 2.258109655260145, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.836077690124512, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8724208474159241, "num_tokens": 677415005.0, "step": 17751 }, { "epoch": 2.2582368655387355, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.709230422973633, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8555565476417542, "num_tokens": 677457908.0, "step": 17752 }, { "epoch": 2.258364075817326, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.731846809387207, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8630651235580444, "num_tokens": 677492759.0, "step": 17753 }, { "epoch": 2.2584912860959165, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.79487419128418, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.870657205581665, "num_tokens": 677535771.0, "step": 17754 }, { "epoch": 2.258618496374507, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.765101432800293, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8856565952301025, "num_tokens": 677575325.0, "step": 17755 }, { "epoch": 2.2587457066530976, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.761696815490723, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8734387755393982, "num_tokens": 677614132.0, "step": 17756 }, { "epoch": 2.258872916931688, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.773914337158203, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8678448796272278, "num_tokens": 677645657.0, "step": 17757 }, { "epoch": 2.2590001272102787, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.732893943786621, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8809404969215393, "num_tokens": 677686238.0, "step": 17758 }, { "epoch": 2.259127337488869, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.696672439575195, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8775322437286377, "num_tokens": 677724080.0, "step": 17759 }, { "epoch": 2.2592545477674597, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.813587188720703, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8733019828796387, "num_tokens": 677765768.0, "step": 17760 }, { "epoch": 2.2593817580460502, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.711464881896973, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8736439347267151, "num_tokens": 677803937.0, "step": 17761 }, { "epoch": 2.2595089683246408, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.71904468536377, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8840307593345642, "num_tokens": 677839685.0, "step": 17762 }, { "epoch": 2.2596361786032313, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.766502380371094, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8584198951721191, "num_tokens": 677880411.0, "step": 17763 }, { "epoch": 2.259763388881822, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.744569778442383, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8813866376876831, "num_tokens": 677921177.0, "step": 17764 }, { "epoch": 2.2598905991604123, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.716252326965332, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8838434815406799, "num_tokens": 677956393.0, "step": 17765 }, { "epoch": 2.2600178094390024, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.725686073303223, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8710196018218994, "num_tokens": 677988364.0, "step": 17766 }, { "epoch": 2.2601450197175934, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.794722557067871, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8807138204574585, "num_tokens": 678028093.0, "step": 17767 }, { "epoch": 2.2602722299961835, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.697498321533203, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8748631477355957, "num_tokens": 678068174.0, "step": 17768 }, { "epoch": 2.260399440274774, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.727866172790527, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8745025992393494, "num_tokens": 678105784.0, "step": 17769 }, { "epoch": 2.2605266505533645, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.719079971313477, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8845766186714172, "num_tokens": 678142873.0, "step": 17770 }, { "epoch": 2.260653860831955, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.792430877685547, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8706934452056885, "num_tokens": 678181527.0, "step": 17771 }, { "epoch": 2.2607810711105456, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.715270042419434, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8760020732879639, "num_tokens": 678221225.0, "step": 17772 }, { "epoch": 2.260908281389136, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.783326148986816, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8819379210472107, "num_tokens": 678254767.0, "step": 17773 }, { "epoch": 2.2610354916677267, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.76714038848877, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8757179975509644, "num_tokens": 678290830.0, "step": 17774 }, { "epoch": 2.261162701946317, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.745413780212402, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.884253203868866, "num_tokens": 678327102.0, "step": 17775 }, { "epoch": 2.2612899122249077, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.738943099975586, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8864270448684692, "num_tokens": 678364872.0, "step": 17776 }, { "epoch": 2.2614171225034982, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.783740043640137, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8673015236854553, "num_tokens": 678399233.0, "step": 17777 }, { "epoch": 2.2615443327820888, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.691999435424805, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8814052939414978, "num_tokens": 678434577.0, "step": 17778 }, { "epoch": 2.2616715430606793, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.882043838500977, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8796674609184265, "num_tokens": 678466835.0, "step": 17779 }, { "epoch": 2.26179875333927, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.669610023498535, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.8884351253509521, "num_tokens": 678506817.0, "step": 17780 }, { "epoch": 2.2619259636178604, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.832202911376953, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8747425079345703, "num_tokens": 678543734.0, "step": 17781 }, { "epoch": 2.262053173896451, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.82703685760498, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8699039220809937, "num_tokens": 678585066.0, "step": 17782 }, { "epoch": 2.2621803841750414, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.820648193359375, "learning_rate": 1e-06, "loss": 0.3458, "mean_token_accuracy": 0.8884737491607666, "num_tokens": 678614951.0, "step": 17783 }, { "epoch": 2.262307594453632, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.818414688110352, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8764635920524597, "num_tokens": 678649671.0, "step": 17784 }, { "epoch": 2.2624348047322225, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.803622245788574, "learning_rate": 1e-06, "loss": 0.3456, "mean_token_accuracy": 0.8902623653411865, "num_tokens": 678693418.0, "step": 17785 }, { "epoch": 2.262562015010813, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.757316589355469, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8839695453643799, "num_tokens": 678730891.0, "step": 17786 }, { "epoch": 2.2626892252894035, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.823562622070312, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8823277950286865, "num_tokens": 678766369.0, "step": 17787 }, { "epoch": 2.262816435567994, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.765387535095215, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8630266785621643, "num_tokens": 678809118.0, "step": 17788 }, { "epoch": 2.2629436458465846, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.78104019165039, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8724843263626099, "num_tokens": 678850057.0, "step": 17789 }, { "epoch": 2.263070856125175, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.796509742736816, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8719445466995239, "num_tokens": 678890253.0, "step": 17790 }, { "epoch": 2.263198066403765, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.754308700561523, "learning_rate": 1e-06, "loss": 0.352, "mean_token_accuracy": 0.8888720870018005, "num_tokens": 678922535.0, "step": 17791 }, { "epoch": 2.263325276682356, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.828902244567871, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8730112910270691, "num_tokens": 678956136.0, "step": 17792 }, { "epoch": 2.2634524869609463, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.73759651184082, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8777111172676086, "num_tokens": 678990864.0, "step": 17793 }, { "epoch": 2.263579697239537, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.759329795837402, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8865442872047424, "num_tokens": 679028345.0, "step": 17794 }, { "epoch": 2.2637069075181273, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.769624710083008, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8616728782653809, "num_tokens": 679073679.0, "step": 17795 }, { "epoch": 2.263834117796718, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.672568321228027, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8729362487792969, "num_tokens": 679109274.0, "step": 17796 }, { "epoch": 2.2639613280753084, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.771113395690918, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8761976361274719, "num_tokens": 679141644.0, "step": 17797 }, { "epoch": 2.264088538353899, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.750651359558105, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8584696054458618, "num_tokens": 679184351.0, "step": 17798 }, { "epoch": 2.2642157486324894, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.728100776672363, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.870155930519104, "num_tokens": 679222509.0, "step": 17799 }, { "epoch": 2.26434295891108, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.767332077026367, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.858564019203186, "num_tokens": 679263362.0, "step": 17800 }, { "epoch": 2.2644701691896705, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.756132125854492, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8628684282302856, "num_tokens": 679307859.0, "step": 17801 }, { "epoch": 2.264597379468261, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.706437110900879, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8844216465950012, "num_tokens": 679344801.0, "step": 17802 }, { "epoch": 2.2647245897468515, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.759690284729004, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8660590648651123, "num_tokens": 679382795.0, "step": 17803 }, { "epoch": 2.264851800025442, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.804883003234863, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8712787628173828, "num_tokens": 679425282.0, "step": 17804 }, { "epoch": 2.2649790103040326, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.718567848205566, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8606129288673401, "num_tokens": 679466881.0, "step": 17805 }, { "epoch": 2.265106220582623, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.791446685791016, "learning_rate": 1e-06, "loss": 0.3542, "mean_token_accuracy": 0.885979950428009, "num_tokens": 679509268.0, "step": 17806 }, { "epoch": 2.2652334308612136, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.726834297180176, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8717396259307861, "num_tokens": 679545082.0, "step": 17807 }, { "epoch": 2.265360641139804, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.72916316986084, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8762445449829102, "num_tokens": 679581795.0, "step": 17808 }, { "epoch": 2.2654878514183947, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.77060317993164, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8723621368408203, "num_tokens": 679618978.0, "step": 17809 }, { "epoch": 2.2656150616969852, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.753616333007812, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8770556449890137, "num_tokens": 679655002.0, "step": 17810 }, { "epoch": 2.2657422719755758, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.791733741760254, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8877131938934326, "num_tokens": 679691143.0, "step": 17811 }, { "epoch": 2.2658694822541663, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.757433891296387, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8760988116264343, "num_tokens": 679729168.0, "step": 17812 }, { "epoch": 2.265996692532757, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.737683296203613, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8727961778640747, "num_tokens": 679773385.0, "step": 17813 }, { "epoch": 2.2661239028113473, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.807657241821289, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8853169083595276, "num_tokens": 679813884.0, "step": 17814 }, { "epoch": 2.266251113089938, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.75684928894043, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8750174045562744, "num_tokens": 679850834.0, "step": 17815 }, { "epoch": 2.266378323368528, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.787471771240234, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.879929780960083, "num_tokens": 679890210.0, "step": 17816 }, { "epoch": 2.266505533647119, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.816445350646973, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8823074698448181, "num_tokens": 679926252.0, "step": 17817 }, { "epoch": 2.266632743925709, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.784141540527344, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8714168667793274, "num_tokens": 679959626.0, "step": 17818 }, { "epoch": 2.2667599542042995, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.708165168762207, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8685896396636963, "num_tokens": 679999973.0, "step": 17819 }, { "epoch": 2.26688716448289, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.778289794921875, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8678770065307617, "num_tokens": 680041081.0, "step": 17820 }, { "epoch": 2.2670143747614806, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.695801734924316, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8787134885787964, "num_tokens": 680078953.0, "step": 17821 }, { "epoch": 2.267141585040071, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.812314987182617, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8829085826873779, "num_tokens": 680112309.0, "step": 17822 }, { "epoch": 2.2672687953186617, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.71491527557373, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8698093891143799, "num_tokens": 680148607.0, "step": 17823 }, { "epoch": 2.267396005597252, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.820297241210938, "learning_rate": 1e-06, "loss": 0.3444, "mean_token_accuracy": 0.8920063972473145, "num_tokens": 680189617.0, "step": 17824 }, { "epoch": 2.2675232158758427, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.726731300354004, "learning_rate": 1e-06, "loss": 0.3461, "mean_token_accuracy": 0.8897299766540527, "num_tokens": 680227940.0, "step": 17825 }, { "epoch": 2.2676504261544332, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.800374984741211, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8871685862541199, "num_tokens": 680263473.0, "step": 17826 }, { "epoch": 2.2677776364330238, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.798796653747559, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8666924834251404, "num_tokens": 680297895.0, "step": 17827 }, { "epoch": 2.2679048467116143, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.849002838134766, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8764092922210693, "num_tokens": 680337658.0, "step": 17828 }, { "epoch": 2.268032056990205, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.762027740478516, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.86846923828125, "num_tokens": 680375912.0, "step": 17829 }, { "epoch": 2.2681592672687954, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.74532413482666, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8860592842102051, "num_tokens": 680415231.0, "step": 17830 }, { "epoch": 2.268286477547386, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.712029457092285, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8696452379226685, "num_tokens": 680460037.0, "step": 17831 }, { "epoch": 2.2684136878259764, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.74941349029541, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8524947166442871, "num_tokens": 680505521.0, "step": 17832 }, { "epoch": 2.268540898104567, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.715277671813965, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8879713416099548, "num_tokens": 680547682.0, "step": 17833 }, { "epoch": 2.2686681083831575, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.791620254516602, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8813819289207458, "num_tokens": 680587510.0, "step": 17834 }, { "epoch": 2.268795318661748, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.755457878112793, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8787088990211487, "num_tokens": 680624233.0, "step": 17835 }, { "epoch": 2.2689225289403385, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.758859634399414, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8699591755867004, "num_tokens": 680666091.0, "step": 17836 }, { "epoch": 2.269049739218929, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.793859481811523, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.880570650100708, "num_tokens": 680703062.0, "step": 17837 }, { "epoch": 2.2691769494975196, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.825115203857422, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8877339363098145, "num_tokens": 680735966.0, "step": 17838 }, { "epoch": 2.2693041597761097, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.80345630645752, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8742980360984802, "num_tokens": 680780010.0, "step": 17839 }, { "epoch": 2.2694313700547006, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.753083229064941, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.874354898929596, "num_tokens": 680819516.0, "step": 17840 }, { "epoch": 2.2695585803332907, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.889616012573242, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8741940259933472, "num_tokens": 680861374.0, "step": 17841 }, { "epoch": 2.2696857906118812, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.802011489868164, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8669666647911072, "num_tokens": 680902447.0, "step": 17842 }, { "epoch": 2.2698130008904718, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.751205444335938, "learning_rate": 1e-06, "loss": 0.3398, "mean_token_accuracy": 0.8900737166404724, "num_tokens": 680942376.0, "step": 17843 }, { "epoch": 2.2699402111690623, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.802913665771484, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8760706782341003, "num_tokens": 680977060.0, "step": 17844 }, { "epoch": 2.270067421447653, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.826400756835938, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8597322702407837, "num_tokens": 681014504.0, "step": 17845 }, { "epoch": 2.2701946317262434, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.828580856323242, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8582782745361328, "num_tokens": 681053919.0, "step": 17846 }, { "epoch": 2.270321842004834, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.79119873046875, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8748083114624023, "num_tokens": 681090883.0, "step": 17847 }, { "epoch": 2.2704490522834244, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.799949645996094, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8883401155471802, "num_tokens": 681128819.0, "step": 17848 }, { "epoch": 2.270576262562015, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.753540992736816, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8655584454536438, "num_tokens": 681171434.0, "step": 17849 }, { "epoch": 2.2707034728406055, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.811222076416016, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8845299482345581, "num_tokens": 681206297.0, "step": 17850 }, { "epoch": 2.270830683119196, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.76080322265625, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8648176193237305, "num_tokens": 681242521.0, "step": 17851 }, { "epoch": 2.2709578933977865, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.772050857543945, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8850531578063965, "num_tokens": 681282474.0, "step": 17852 }, { "epoch": 2.271085103676377, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.714308738708496, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8708187937736511, "num_tokens": 681317280.0, "step": 17853 }, { "epoch": 2.2712123139549676, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.840483665466309, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8730082511901855, "num_tokens": 681349155.0, "step": 17854 }, { "epoch": 2.271339524233558, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.806168556213379, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8748313188552856, "num_tokens": 681380863.0, "step": 17855 }, { "epoch": 2.2714667345121486, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.73654842376709, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8786100745201111, "num_tokens": 681421970.0, "step": 17856 }, { "epoch": 2.271593944790739, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.736640930175781, "learning_rate": 1e-06, "loss": 0.3436, "mean_token_accuracy": 0.8891009092330933, "num_tokens": 681453511.0, "step": 17857 }, { "epoch": 2.2717211550693297, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.743563652038574, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8617210984230042, "num_tokens": 681496293.0, "step": 17858 }, { "epoch": 2.2718483653479202, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.75212574005127, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8675333261489868, "num_tokens": 681537074.0, "step": 17859 }, { "epoch": 2.2719755756265108, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.78550910949707, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8673503398895264, "num_tokens": 681578911.0, "step": 17860 }, { "epoch": 2.2721027859051013, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.782366752624512, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8710246682167053, "num_tokens": 681622134.0, "step": 17861 }, { "epoch": 2.272229996183692, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.814748764038086, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8765867352485657, "num_tokens": 681655919.0, "step": 17862 }, { "epoch": 2.2723572064622823, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.706121444702148, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8664196133613586, "num_tokens": 681696539.0, "step": 17863 }, { "epoch": 2.2724844167408724, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.794180870056152, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8729745149612427, "num_tokens": 681734630.0, "step": 17864 }, { "epoch": 2.2726116270194634, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.743827819824219, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.862563967704773, "num_tokens": 681779669.0, "step": 17865 }, { "epoch": 2.2727388372980535, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.818504333496094, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8719635009765625, "num_tokens": 681821920.0, "step": 17866 }, { "epoch": 2.272866047576644, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.732649803161621, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8725681900978088, "num_tokens": 681859239.0, "step": 17867 }, { "epoch": 2.2729932578552345, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.729011535644531, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.862082839012146, "num_tokens": 681896569.0, "step": 17868 }, { "epoch": 2.273120468133825, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.853636741638184, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8892470598220825, "num_tokens": 681932549.0, "step": 17869 }, { "epoch": 2.2732476784124156, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.753387451171875, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8702635765075684, "num_tokens": 681973981.0, "step": 17870 }, { "epoch": 2.273374888691006, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.75425910949707, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8525453209877014, "num_tokens": 682009512.0, "step": 17871 }, { "epoch": 2.2735020989695967, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.818158149719238, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8725273609161377, "num_tokens": 682048058.0, "step": 17872 }, { "epoch": 2.273629309248187, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.69351577758789, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8662276268005371, "num_tokens": 682084181.0, "step": 17873 }, { "epoch": 2.2737565195267777, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83968448638916, "learning_rate": 1e-06, "loss": 0.3405, "mean_token_accuracy": 0.8934506177902222, "num_tokens": 682119613.0, "step": 17874 }, { "epoch": 2.2738837298053682, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.715767860412598, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8779841661453247, "num_tokens": 682159501.0, "step": 17875 }, { "epoch": 2.2740109400839588, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.776054382324219, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8824470639228821, "num_tokens": 682197323.0, "step": 17876 }, { "epoch": 2.2741381503625493, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.768318176269531, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8754351139068604, "num_tokens": 682231150.0, "step": 17877 }, { "epoch": 2.27426536064114, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.784225463867188, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8760441541671753, "num_tokens": 682269351.0, "step": 17878 }, { "epoch": 2.2743925709197303, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.866496086120605, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8853323459625244, "num_tokens": 682303633.0, "step": 17879 }, { "epoch": 2.274519781198321, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.757661819458008, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8730502724647522, "num_tokens": 682340890.0, "step": 17880 }, { "epoch": 2.2746469914769114, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.758426666259766, "learning_rate": 1e-06, "loss": 0.3485, "mean_token_accuracy": 0.8889222741127014, "num_tokens": 682374253.0, "step": 17881 }, { "epoch": 2.274774201755502, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.856395721435547, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8741618990898132, "num_tokens": 682412675.0, "step": 17882 }, { "epoch": 2.2749014120340925, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.868878364562988, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8857660293579102, "num_tokens": 682447864.0, "step": 17883 }, { "epoch": 2.275028622312683, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.755149841308594, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8748885989189148, "num_tokens": 682486445.0, "step": 17884 }, { "epoch": 2.2751558325912735, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.800626754760742, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8739488124847412, "num_tokens": 682518173.0, "step": 17885 }, { "epoch": 2.275283042869864, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.741019248962402, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.87358558177948, "num_tokens": 682558627.0, "step": 17886 }, { "epoch": 2.2754102531484546, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.712543487548828, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8792881965637207, "num_tokens": 682600821.0, "step": 17887 }, { "epoch": 2.275537463427045, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.822181701660156, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8622106313705444, "num_tokens": 682643947.0, "step": 17888 }, { "epoch": 2.275664673705635, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.785017013549805, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8695284128189087, "num_tokens": 682683045.0, "step": 17889 }, { "epoch": 2.275791883984226, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.756250381469727, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8753992915153503, "num_tokens": 682721617.0, "step": 17890 }, { "epoch": 2.2759190942628162, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.768051147460938, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8680822849273682, "num_tokens": 682758339.0, "step": 17891 }, { "epoch": 2.2760463045414068, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.767545700073242, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8793264031410217, "num_tokens": 682800501.0, "step": 17892 }, { "epoch": 2.2761735148199973, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.794313430786133, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8760664463043213, "num_tokens": 682833231.0, "step": 17893 }, { "epoch": 2.276300725098588, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.763012886047363, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.879761815071106, "num_tokens": 682873945.0, "step": 17894 }, { "epoch": 2.2764279353771784, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.761613845825195, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8733512163162231, "num_tokens": 682918329.0, "step": 17895 }, { "epoch": 2.276555145655769, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.7606782913208, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8800661563873291, "num_tokens": 682956277.0, "step": 17896 }, { "epoch": 2.2766823559343594, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.780253410339355, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8785628080368042, "num_tokens": 682990648.0, "step": 17897 }, { "epoch": 2.27680956621295, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.722594261169434, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8891597986221313, "num_tokens": 683029046.0, "step": 17898 }, { "epoch": 2.2769367764915405, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.760322570800781, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8751093745231628, "num_tokens": 683064541.0, "step": 17899 }, { "epoch": 2.277063986770131, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.737330436706543, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8744570016860962, "num_tokens": 683105015.0, "step": 17900 }, { "epoch": 2.2771911970487215, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.777850151062012, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8685015439987183, "num_tokens": 683142938.0, "step": 17901 }, { "epoch": 2.277318407327312, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.735241889953613, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.88800048828125, "num_tokens": 683181593.0, "step": 17902 }, { "epoch": 2.2774456176059026, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.770613670349121, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.863395631313324, "num_tokens": 683218641.0, "step": 17903 }, { "epoch": 2.277572827884493, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.758564949035645, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8591687679290771, "num_tokens": 683261479.0, "step": 17904 }, { "epoch": 2.2777000381630836, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.765704154968262, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8491538166999817, "num_tokens": 683300602.0, "step": 17905 }, { "epoch": 2.277827248441674, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.766337394714355, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8727604746818542, "num_tokens": 683337012.0, "step": 17906 }, { "epoch": 2.2779544587202647, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.727180480957031, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8823148012161255, "num_tokens": 683368949.0, "step": 17907 }, { "epoch": 2.2780816689988552, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.781819343566895, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8721997737884521, "num_tokens": 683408741.0, "step": 17908 }, { "epoch": 2.2782088792774458, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.710824966430664, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8643631339073181, "num_tokens": 683447843.0, "step": 17909 }, { "epoch": 2.2783360895560363, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.790475845336914, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.871118426322937, "num_tokens": 683489695.0, "step": 17910 }, { "epoch": 2.278463299834627, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.71922779083252, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8724846839904785, "num_tokens": 683528381.0, "step": 17911 }, { "epoch": 2.2785905101132173, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.827157020568848, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8756864070892334, "num_tokens": 683567261.0, "step": 17912 }, { "epoch": 2.278717720391808, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.725172996520996, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8615727424621582, "num_tokens": 683601083.0, "step": 17913 }, { "epoch": 2.278844930670398, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.77042293548584, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8676717281341553, "num_tokens": 683643494.0, "step": 17914 }, { "epoch": 2.278972140948989, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.745504379272461, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8703280687332153, "num_tokens": 683684267.0, "step": 17915 }, { "epoch": 2.279099351227579, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.716073036193848, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8628787994384766, "num_tokens": 683723292.0, "step": 17916 }, { "epoch": 2.2792265615061695, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.733453750610352, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.884567379951477, "num_tokens": 683761378.0, "step": 17917 }, { "epoch": 2.27935377178476, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.730382919311523, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8661231994628906, "num_tokens": 683794838.0, "step": 17918 }, { "epoch": 2.2794809820633506, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.805584907531738, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8559929728507996, "num_tokens": 683832307.0, "step": 17919 }, { "epoch": 2.279608192341941, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.658062934875488, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8670284748077393, "num_tokens": 683871848.0, "step": 17920 }, { "epoch": 2.2797354026205316, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.783546447753906, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8779277801513672, "num_tokens": 683915086.0, "step": 17921 }, { "epoch": 2.279862612899122, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.793704986572266, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8670388460159302, "num_tokens": 683945411.0, "step": 17922 }, { "epoch": 2.2799898231777127, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.793508529663086, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8740885257720947, "num_tokens": 683979664.0, "step": 17923 }, { "epoch": 2.2801170334563032, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.760912895202637, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8641811013221741, "num_tokens": 684025526.0, "step": 17924 }, { "epoch": 2.2802442437348938, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.825630187988281, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.868003249168396, "num_tokens": 684056992.0, "step": 17925 }, { "epoch": 2.2803714540134843, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.806694984436035, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8774741888046265, "num_tokens": 684094515.0, "step": 17926 }, { "epoch": 2.280498664292075, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.76191520690918, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8807801008224487, "num_tokens": 684132294.0, "step": 17927 }, { "epoch": 2.2806258745706653, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.771828651428223, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8773282766342163, "num_tokens": 684164010.0, "step": 17928 }, { "epoch": 2.280753084849256, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.843351364135742, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8810625076293945, "num_tokens": 684202077.0, "step": 17929 }, { "epoch": 2.2808802951278464, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.728428840637207, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.847394585609436, "num_tokens": 684240799.0, "step": 17930 }, { "epoch": 2.281007505406437, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.75638484954834, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8708642721176147, "num_tokens": 684284080.0, "step": 17931 }, { "epoch": 2.2811347156850275, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.804549217224121, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8571525812149048, "num_tokens": 684317915.0, "step": 17932 }, { "epoch": 2.281261925963618, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.751240730285645, "learning_rate": 1e-06, "loss": 0.2841, "mean_token_accuracy": 0.9105879664421082, "num_tokens": 684354665.0, "step": 17933 }, { "epoch": 2.2813891362422085, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.811074256896973, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8727572560310364, "num_tokens": 684392866.0, "step": 17934 }, { "epoch": 2.281516346520799, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.838140487670898, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8776452541351318, "num_tokens": 684428422.0, "step": 17935 }, { "epoch": 2.2816435567993896, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.81643009185791, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8613787889480591, "num_tokens": 684467129.0, "step": 17936 }, { "epoch": 2.2817707670779797, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.807132720947266, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.867438018321991, "num_tokens": 684500093.0, "step": 17937 }, { "epoch": 2.2818979773565706, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.839823722839355, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8647085428237915, "num_tokens": 684541422.0, "step": 17938 }, { "epoch": 2.2820251876351607, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.727407455444336, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8700289726257324, "num_tokens": 684579164.0, "step": 17939 }, { "epoch": 2.2821523979137512, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.786334037780762, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8581205606460571, "num_tokens": 684618617.0, "step": 17940 }, { "epoch": 2.2822796081923418, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.777299880981445, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8596678376197815, "num_tokens": 684655525.0, "step": 17941 }, { "epoch": 2.2824068184709323, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.79568862915039, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8727372884750366, "num_tokens": 684694525.0, "step": 17942 }, { "epoch": 2.282534028749523, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.755297660827637, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8740164041519165, "num_tokens": 684733247.0, "step": 17943 }, { "epoch": 2.2826612390281134, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.78434944152832, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8864407539367676, "num_tokens": 684763642.0, "step": 17944 }, { "epoch": 2.282788449306704, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.75830364227295, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8798506259918213, "num_tokens": 684806164.0, "step": 17945 }, { "epoch": 2.2829156595852944, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.7766752243042, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8716689348220825, "num_tokens": 684844193.0, "step": 17946 }, { "epoch": 2.283042869863885, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.643036842346191, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8786265850067139, "num_tokens": 684882103.0, "step": 17947 }, { "epoch": 2.2831700801424755, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.777114868164062, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.887736439704895, "num_tokens": 684917667.0, "step": 17948 }, { "epoch": 2.283297290421066, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.695525169372559, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8776874542236328, "num_tokens": 684954007.0, "step": 17949 }, { "epoch": 2.2834245006996565, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.773150444030762, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.868907630443573, "num_tokens": 684994693.0, "step": 17950 }, { "epoch": 2.283551710978247, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.785543441772461, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.860451340675354, "num_tokens": 685032713.0, "step": 17951 }, { "epoch": 2.2836789212568376, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.76852798461914, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8678092360496521, "num_tokens": 685066597.0, "step": 17952 }, { "epoch": 2.283806131535428, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.772211074829102, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8682494163513184, "num_tokens": 685106243.0, "step": 17953 }, { "epoch": 2.2839333418140186, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.769205093383789, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8645789623260498, "num_tokens": 685145177.0, "step": 17954 }, { "epoch": 2.284060552092609, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.763395309448242, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8796585202217102, "num_tokens": 685189130.0, "step": 17955 }, { "epoch": 2.2841877623711997, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.764921188354492, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8782734870910645, "num_tokens": 685227536.0, "step": 17956 }, { "epoch": 2.28431497264979, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.781937599182129, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8784745335578918, "num_tokens": 685267213.0, "step": 17957 }, { "epoch": 2.2844421829283807, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.829911231994629, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8711514472961426, "num_tokens": 685301115.0, "step": 17958 }, { "epoch": 2.2845693932069713, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.876105308532715, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.878662109375, "num_tokens": 685338497.0, "step": 17959 }, { "epoch": 2.284696603485562, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.69480037689209, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8632891774177551, "num_tokens": 685374983.0, "step": 17960 }, { "epoch": 2.2848238137641523, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.792510032653809, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8792100548744202, "num_tokens": 685408859.0, "step": 17961 }, { "epoch": 2.2849510240427424, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.755219459533691, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8698670864105225, "num_tokens": 685445103.0, "step": 17962 }, { "epoch": 2.2850782343213334, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.777698516845703, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8685993552207947, "num_tokens": 685484657.0, "step": 17963 }, { "epoch": 2.2852054445999235, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.807354927062988, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8817269802093506, "num_tokens": 685524034.0, "step": 17964 }, { "epoch": 2.285332654878514, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.838226318359375, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8789218664169312, "num_tokens": 685560368.0, "step": 17965 }, { "epoch": 2.2854598651571045, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.856761932373047, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8702198266983032, "num_tokens": 685598192.0, "step": 17966 }, { "epoch": 2.285587075435695, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.870766639709473, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8758484125137329, "num_tokens": 685636300.0, "step": 17967 }, { "epoch": 2.2857142857142856, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.710092544555664, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8710127472877502, "num_tokens": 685671457.0, "step": 17968 }, { "epoch": 2.285841495992876, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.746085166931152, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8635580539703369, "num_tokens": 685709687.0, "step": 17969 }, { "epoch": 2.2859687062714666, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.820826530456543, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8515862226486206, "num_tokens": 685747854.0, "step": 17970 }, { "epoch": 2.286095916550057, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.835186958312988, "learning_rate": 1e-06, "loss": 0.3426, "mean_token_accuracy": 0.8930907249450684, "num_tokens": 685785215.0, "step": 17971 }, { "epoch": 2.2862231268286477, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.73556137084961, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8774734735488892, "num_tokens": 685818871.0, "step": 17972 }, { "epoch": 2.2863503371072382, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.758562088012695, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8671481609344482, "num_tokens": 685859089.0, "step": 17973 }, { "epoch": 2.2864775473858288, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.743791580200195, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8643785715103149, "num_tokens": 685900346.0, "step": 17974 }, { "epoch": 2.2866047576644193, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.777334213256836, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8726785778999329, "num_tokens": 685938186.0, "step": 17975 }, { "epoch": 2.28673196794301, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.73680591583252, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8803984522819519, "num_tokens": 685974911.0, "step": 17976 }, { "epoch": 2.2868591782216003, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.829343795776367, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8750433921813965, "num_tokens": 686010430.0, "step": 17977 }, { "epoch": 2.286986388500191, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.684852600097656, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.886843204498291, "num_tokens": 686042727.0, "step": 17978 }, { "epoch": 2.2871135987787814, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.81541633605957, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8748772740364075, "num_tokens": 686078819.0, "step": 17979 }, { "epoch": 2.287240809057372, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.748725891113281, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8757954835891724, "num_tokens": 686117760.0, "step": 17980 }, { "epoch": 2.2873680193359625, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.770502090454102, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8645797967910767, "num_tokens": 686152992.0, "step": 17981 }, { "epoch": 2.287495229614553, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.758980751037598, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8677659034729004, "num_tokens": 686195353.0, "step": 17982 }, { "epoch": 2.2876224398931435, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.729608535766602, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8692731857299805, "num_tokens": 686232990.0, "step": 17983 }, { "epoch": 2.287749650171734, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.802631378173828, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.877915620803833, "num_tokens": 686265933.0, "step": 17984 }, { "epoch": 2.2878768604503246, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.747236251831055, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8797209858894348, "num_tokens": 686307111.0, "step": 17985 }, { "epoch": 2.288004070728915, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.80371379852295, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8754411935806274, "num_tokens": 686343363.0, "step": 17986 }, { "epoch": 2.288131281007505, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.765104293823242, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8721850514411926, "num_tokens": 686380888.0, "step": 17987 }, { "epoch": 2.288258491286096, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.768013954162598, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.864617109298706, "num_tokens": 686417350.0, "step": 17988 }, { "epoch": 2.2883857015646862, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.78036117553711, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8814413547515869, "num_tokens": 686458658.0, "step": 17989 }, { "epoch": 2.2885129118432768, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.768902778625488, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8768705129623413, "num_tokens": 686493945.0, "step": 17990 }, { "epoch": 2.2886401221218673, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.778389930725098, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8728630542755127, "num_tokens": 686527608.0, "step": 17991 }, { "epoch": 2.288767332400458, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.81911849975586, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8828869462013245, "num_tokens": 686562272.0, "step": 17992 }, { "epoch": 2.2888945426790483, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.719832420349121, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8845233917236328, "num_tokens": 686599937.0, "step": 17993 }, { "epoch": 2.289021752957639, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.709826469421387, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8741532564163208, "num_tokens": 686632103.0, "step": 17994 }, { "epoch": 2.2891489632362294, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.763206481933594, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8788031339645386, "num_tokens": 686662802.0, "step": 17995 }, { "epoch": 2.28927617351482, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.807978630065918, "learning_rate": 1e-06, "loss": 0.3368, "mean_token_accuracy": 0.8939993381500244, "num_tokens": 686699386.0, "step": 17996 }, { "epoch": 2.2894033837934105, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.713114738464355, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8769199252128601, "num_tokens": 686738369.0, "step": 17997 }, { "epoch": 2.289530594072001, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.733320236206055, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8667659759521484, "num_tokens": 686774543.0, "step": 17998 }, { "epoch": 2.2896578043505915, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.745121955871582, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8664501905441284, "num_tokens": 686807048.0, "step": 17999 }, { "epoch": 2.289785014629182, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.744330406188965, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8742051720619202, "num_tokens": 686844678.0, "step": 18000 }, { "epoch": 2.2899122249077726, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.722955703735352, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8725976943969727, "num_tokens": 686877721.0, "step": 18001 }, { "epoch": 2.290039435186363, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.786477088928223, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8667662143707275, "num_tokens": 686918329.0, "step": 18002 }, { "epoch": 2.2901666454649536, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.71498966217041, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8687999844551086, "num_tokens": 686960050.0, "step": 18003 }, { "epoch": 2.290293855743544, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.77638053894043, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8831643462181091, "num_tokens": 686997849.0, "step": 18004 }, { "epoch": 2.2904210660221347, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.74543571472168, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8769268989562988, "num_tokens": 687032363.0, "step": 18005 }, { "epoch": 2.290548276300725, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.736198425292969, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8581153154373169, "num_tokens": 687069606.0, "step": 18006 }, { "epoch": 2.2906754865793157, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.759458541870117, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8642557859420776, "num_tokens": 687112857.0, "step": 18007 }, { "epoch": 2.2908026968579063, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.773747444152832, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8711777329444885, "num_tokens": 687152245.0, "step": 18008 }, { "epoch": 2.290929907136497, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.732898712158203, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8733260631561279, "num_tokens": 687193988.0, "step": 18009 }, { "epoch": 2.2910571174150873, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.743124008178711, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8686959743499756, "num_tokens": 687232375.0, "step": 18010 }, { "epoch": 2.291184327693678, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.741217613220215, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8768041729927063, "num_tokens": 687274566.0, "step": 18011 }, { "epoch": 2.291311537972268, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.806846618652344, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8725644946098328, "num_tokens": 687313231.0, "step": 18012 }, { "epoch": 2.291438748250859, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.718764305114746, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8782221078872681, "num_tokens": 687354502.0, "step": 18013 }, { "epoch": 2.291565958529449, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.789729118347168, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8703620433807373, "num_tokens": 687399160.0, "step": 18014 }, { "epoch": 2.2916931688080395, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.763710975646973, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.863336443901062, "num_tokens": 687438775.0, "step": 18015 }, { "epoch": 2.29182037908663, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.832895278930664, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8772889375686646, "num_tokens": 687471435.0, "step": 18016 }, { "epoch": 2.2919475893652206, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.74038314819336, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.874443531036377, "num_tokens": 687508657.0, "step": 18017 }, { "epoch": 2.292074799643811, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.75462818145752, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8835901021957397, "num_tokens": 687545024.0, "step": 18018 }, { "epoch": 2.2922020099224016, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.79359245300293, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8777384757995605, "num_tokens": 687579432.0, "step": 18019 }, { "epoch": 2.292329220200992, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.819121360778809, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8637629747390747, "num_tokens": 687618378.0, "step": 18020 }, { "epoch": 2.2924564304795827, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.795548439025879, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8722395896911621, "num_tokens": 687657119.0, "step": 18021 }, { "epoch": 2.2925836407581732, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.75875473022461, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8629469871520996, "num_tokens": 687691924.0, "step": 18022 }, { "epoch": 2.2927108510367638, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.856732368469238, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.850954532623291, "num_tokens": 687724221.0, "step": 18023 }, { "epoch": 2.2928380613153543, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.7777681350708, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.875176191329956, "num_tokens": 687765049.0, "step": 18024 }, { "epoch": 2.292965271593945, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.786144256591797, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8644484281539917, "num_tokens": 687810222.0, "step": 18025 }, { "epoch": 2.2930924818725353, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.798277854919434, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8684614300727844, "num_tokens": 687851961.0, "step": 18026 }, { "epoch": 2.293219692151126, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.786017417907715, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.861543595790863, "num_tokens": 687890448.0, "step": 18027 }, { "epoch": 2.2933469024297164, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.703289985656738, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.882175087928772, "num_tokens": 687929509.0, "step": 18028 }, { "epoch": 2.293474112708307, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.876978874206543, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8634700775146484, "num_tokens": 687966814.0, "step": 18029 }, { "epoch": 2.2936013229868975, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.684359550476074, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8803826570510864, "num_tokens": 688006398.0, "step": 18030 }, { "epoch": 2.293728533265488, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.846070289611816, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8716657161712646, "num_tokens": 688044604.0, "step": 18031 }, { "epoch": 2.2938557435440785, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.777332305908203, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8687376379966736, "num_tokens": 688082969.0, "step": 18032 }, { "epoch": 2.293982953822669, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.784233093261719, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8799868226051331, "num_tokens": 688121442.0, "step": 18033 }, { "epoch": 2.2941101641012596, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.727241516113281, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8734685182571411, "num_tokens": 688161096.0, "step": 18034 }, { "epoch": 2.2942373743798496, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.836997032165527, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8733394145965576, "num_tokens": 688200782.0, "step": 18035 }, { "epoch": 2.2943645846584406, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.665648460388184, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8805618286132812, "num_tokens": 688241333.0, "step": 18036 }, { "epoch": 2.2944917949370307, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.852914810180664, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8713302612304688, "num_tokens": 688278297.0, "step": 18037 }, { "epoch": 2.2946190052156212, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.7590970993042, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8495237827301025, "num_tokens": 688318835.0, "step": 18038 }, { "epoch": 2.2947462154942118, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.771871566772461, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8756951093673706, "num_tokens": 688354814.0, "step": 18039 }, { "epoch": 2.2948734257728023, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.792441368103027, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8757659196853638, "num_tokens": 688390008.0, "step": 18040 }, { "epoch": 2.295000636051393, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.77906608581543, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8755344152450562, "num_tokens": 688430938.0, "step": 18041 }, { "epoch": 2.2951278463299833, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.767738342285156, "learning_rate": 1e-06, "loss": 0.5146, "mean_token_accuracy": 0.8425413966178894, "num_tokens": 688466730.0, "step": 18042 }, { "epoch": 2.295255056608574, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.839282989501953, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8717958927154541, "num_tokens": 688506666.0, "step": 18043 }, { "epoch": 2.2953822668871644, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.832176208496094, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8591755032539368, "num_tokens": 688547726.0, "step": 18044 }, { "epoch": 2.295509477165755, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.81493854522705, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8691787719726562, "num_tokens": 688587487.0, "step": 18045 }, { "epoch": 2.2956366874443455, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.790785789489746, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8668231964111328, "num_tokens": 688632454.0, "step": 18046 }, { "epoch": 2.295763897722936, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.860356330871582, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8796616792678833, "num_tokens": 688668489.0, "step": 18047 }, { "epoch": 2.2958911080015265, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.742061614990234, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8761885166168213, "num_tokens": 688707899.0, "step": 18048 }, { "epoch": 2.296018318280117, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.76068115234375, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8783844113349915, "num_tokens": 688747171.0, "step": 18049 }, { "epoch": 2.2961455285587076, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.710729598999023, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8706486225128174, "num_tokens": 688789553.0, "step": 18050 }, { "epoch": 2.296272738837298, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.779017448425293, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8771343231201172, "num_tokens": 688827632.0, "step": 18051 }, { "epoch": 2.2963999491158886, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.769350051879883, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8813226222991943, "num_tokens": 688861505.0, "step": 18052 }, { "epoch": 2.296527159394479, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.803098678588867, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.866847038269043, "num_tokens": 688898270.0, "step": 18053 }, { "epoch": 2.2966543696730697, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.793876647949219, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8738659620285034, "num_tokens": 688937420.0, "step": 18054 }, { "epoch": 2.29678157995166, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.788431167602539, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8795076608657837, "num_tokens": 688972426.0, "step": 18055 }, { "epoch": 2.2969087902302507, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.809792518615723, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8728406429290771, "num_tokens": 689005801.0, "step": 18056 }, { "epoch": 2.2970360005088413, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.724316596984863, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8692062497138977, "num_tokens": 689039999.0, "step": 18057 }, { "epoch": 2.297163210787432, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.759515762329102, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.857820451259613, "num_tokens": 689079404.0, "step": 18058 }, { "epoch": 2.2972904210660223, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.735033988952637, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.850921630859375, "num_tokens": 689116197.0, "step": 18059 }, { "epoch": 2.2974176313446124, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.779523849487305, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8742690086364746, "num_tokens": 689153446.0, "step": 18060 }, { "epoch": 2.2975448416232034, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.786079406738281, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.855604350566864, "num_tokens": 689185310.0, "step": 18061 }, { "epoch": 2.2976720519017935, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.850188255310059, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8836358189582825, "num_tokens": 689224150.0, "step": 18062 }, { "epoch": 2.297799262180384, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.752449035644531, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.862985372543335, "num_tokens": 689261206.0, "step": 18063 }, { "epoch": 2.2979264724589745, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.772565841674805, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.877983570098877, "num_tokens": 689299129.0, "step": 18064 }, { "epoch": 2.298053682737565, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.798564910888672, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8883895874023438, "num_tokens": 689330554.0, "step": 18065 }, { "epoch": 2.2981808930161556, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.740487098693848, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8793191909790039, "num_tokens": 689364084.0, "step": 18066 }, { "epoch": 2.298308103294746, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.734621047973633, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8614345788955688, "num_tokens": 689405254.0, "step": 18067 }, { "epoch": 2.2984353135733366, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.804147720336914, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.863298773765564, "num_tokens": 689441109.0, "step": 18068 }, { "epoch": 2.298562523851927, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.74254035949707, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8732907772064209, "num_tokens": 689474937.0, "step": 18069 }, { "epoch": 2.2986897341305177, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.7562255859375, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8674752116203308, "num_tokens": 689515971.0, "step": 18070 }, { "epoch": 2.298816944409108, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.82950210571289, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8507663011550903, "num_tokens": 689553620.0, "step": 18071 }, { "epoch": 2.2989441546876987, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.827430725097656, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8764762282371521, "num_tokens": 689592157.0, "step": 18072 }, { "epoch": 2.2990713649662893, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.755701065063477, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8856714367866516, "num_tokens": 689626889.0, "step": 18073 }, { "epoch": 2.29919857524488, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.835465431213379, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8850187063217163, "num_tokens": 689666161.0, "step": 18074 }, { "epoch": 2.2993257855234703, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.801810264587402, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8649795055389404, "num_tokens": 689713713.0, "step": 18075 }, { "epoch": 2.299452995802061, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.826379776000977, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8651232123374939, "num_tokens": 689755146.0, "step": 18076 }, { "epoch": 2.2995802060806514, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.852143287658691, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8811894655227661, "num_tokens": 689795925.0, "step": 18077 }, { "epoch": 2.299707416359242, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.811845779418945, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.883897066116333, "num_tokens": 689830673.0, "step": 18078 }, { "epoch": 2.2998346266378324, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.852734565734863, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8728185892105103, "num_tokens": 689863649.0, "step": 18079 }, { "epoch": 2.299961836916423, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.735269546508789, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8723549842834473, "num_tokens": 689901946.0, "step": 18080 }, { "epoch": 2.3000890471950135, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.847953796386719, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8670993447303772, "num_tokens": 689939968.0, "step": 18081 }, { "epoch": 2.300216257473604, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.771184921264648, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8817697167396545, "num_tokens": 689979792.0, "step": 18082 }, { "epoch": 2.3003434677521946, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.807315826416016, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8848903775215149, "num_tokens": 690013974.0, "step": 18083 }, { "epoch": 2.300470678030785, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.743626594543457, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8808178901672363, "num_tokens": 690056540.0, "step": 18084 }, { "epoch": 2.300597888309375, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.767799377441406, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8668338060379028, "num_tokens": 690101032.0, "step": 18085 }, { "epoch": 2.300725098587966, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.813831329345703, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8851369023323059, "num_tokens": 690135417.0, "step": 18086 }, { "epoch": 2.3008523088665562, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.786160469055176, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8746193647384644, "num_tokens": 690178552.0, "step": 18087 }, { "epoch": 2.3009795191451468, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.797035217285156, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8703553080558777, "num_tokens": 690221928.0, "step": 18088 }, { "epoch": 2.3011067294237373, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.796995162963867, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8792893290519714, "num_tokens": 690257040.0, "step": 18089 }, { "epoch": 2.301233939702328, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.744521141052246, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8730049133300781, "num_tokens": 690302720.0, "step": 18090 }, { "epoch": 2.3013611499809183, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.848740577697754, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8689462542533875, "num_tokens": 690337042.0, "step": 18091 }, { "epoch": 2.301488360259509, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.783549308776855, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8789350986480713, "num_tokens": 690380247.0, "step": 18092 }, { "epoch": 2.3016155705380994, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.76280689239502, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8776330947875977, "num_tokens": 690425361.0, "step": 18093 }, { "epoch": 2.30174278081669, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.801925659179688, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8679555654525757, "num_tokens": 690465137.0, "step": 18094 }, { "epoch": 2.3018699910952805, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.779207229614258, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8704506754875183, "num_tokens": 690501437.0, "step": 18095 }, { "epoch": 2.301997201373871, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.786900520324707, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.867757260799408, "num_tokens": 690544134.0, "step": 18096 }, { "epoch": 2.3021244116524615, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.754471778869629, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.881680428981781, "num_tokens": 690581062.0, "step": 18097 }, { "epoch": 2.302251621931052, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.873383522033691, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8774986267089844, "num_tokens": 690615145.0, "step": 18098 }, { "epoch": 2.3023788322096426, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.781865119934082, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.878899872303009, "num_tokens": 690655483.0, "step": 18099 }, { "epoch": 2.302506042488233, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.769708633422852, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8702260255813599, "num_tokens": 690694133.0, "step": 18100 }, { "epoch": 2.3026332527668236, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.799166679382324, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8512595295906067, "num_tokens": 690729939.0, "step": 18101 }, { "epoch": 2.302760463045414, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.798996925354004, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8799604177474976, "num_tokens": 690770139.0, "step": 18102 }, { "epoch": 2.3028876733240047, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.769062042236328, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8744148015975952, "num_tokens": 690811379.0, "step": 18103 }, { "epoch": 2.303014883602595, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.823599815368652, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8691459894180298, "num_tokens": 690844194.0, "step": 18104 }, { "epoch": 2.3031420938811857, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.788053512573242, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8666154146194458, "num_tokens": 690882129.0, "step": 18105 }, { "epoch": 2.3032693041597763, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.807295799255371, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8856422901153564, "num_tokens": 690915568.0, "step": 18106 }, { "epoch": 2.303396514438367, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.76732063293457, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8712636232376099, "num_tokens": 690953412.0, "step": 18107 }, { "epoch": 2.3035237247169573, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.818065643310547, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8703337907791138, "num_tokens": 690990047.0, "step": 18108 }, { "epoch": 2.303650934995548, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.836642265319824, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.878138542175293, "num_tokens": 691031010.0, "step": 18109 }, { "epoch": 2.303778145274138, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.80625057220459, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8746910095214844, "num_tokens": 691065830.0, "step": 18110 }, { "epoch": 2.303905355552729, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.867254257202148, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8844451904296875, "num_tokens": 691101418.0, "step": 18111 }, { "epoch": 2.304032565831319, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.786323547363281, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8589376211166382, "num_tokens": 691139447.0, "step": 18112 }, { "epoch": 2.3041597761099095, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.88601016998291, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8549057245254517, "num_tokens": 691168809.0, "step": 18113 }, { "epoch": 2.3042869863885, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.799516677856445, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.878799319267273, "num_tokens": 691200420.0, "step": 18114 }, { "epoch": 2.3044141966670906, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.825323104858398, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8670865893363953, "num_tokens": 691238587.0, "step": 18115 }, { "epoch": 2.304541406945681, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.73930549621582, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8711007833480835, "num_tokens": 691277137.0, "step": 18116 }, { "epoch": 2.3046686172242716, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.835927963256836, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8716020584106445, "num_tokens": 691319564.0, "step": 18117 }, { "epoch": 2.304795827502862, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.784895896911621, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8713189363479614, "num_tokens": 691359652.0, "step": 18118 }, { "epoch": 2.3049230377814527, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.775548934936523, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8727612495422363, "num_tokens": 691399705.0, "step": 18119 }, { "epoch": 2.305050248060043, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.818718910217285, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.866828441619873, "num_tokens": 691439447.0, "step": 18120 }, { "epoch": 2.3051774583386337, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.849355697631836, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8614746332168579, "num_tokens": 691477453.0, "step": 18121 }, { "epoch": 2.3053046686172243, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.694792747497559, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8726152181625366, "num_tokens": 691517766.0, "step": 18122 }, { "epoch": 2.305431878895815, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.979522705078125, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8718980550765991, "num_tokens": 691553956.0, "step": 18123 }, { "epoch": 2.3055590891744053, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.804640769958496, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.871748685836792, "num_tokens": 691592522.0, "step": 18124 }, { "epoch": 2.305686299452996, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.797881126403809, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8720664381980896, "num_tokens": 691629202.0, "step": 18125 }, { "epoch": 2.3058135097315864, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.8617525100708, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8754209280014038, "num_tokens": 691665324.0, "step": 18126 }, { "epoch": 2.305940720010177, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.80989933013916, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8827717304229736, "num_tokens": 691704223.0, "step": 18127 }, { "epoch": 2.3060679302887674, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.799734115600586, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8696897029876709, "num_tokens": 691745061.0, "step": 18128 }, { "epoch": 2.306195140567358, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.807311058044434, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8844234943389893, "num_tokens": 691783630.0, "step": 18129 }, { "epoch": 2.3063223508459485, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.74491024017334, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8711274862289429, "num_tokens": 691818283.0, "step": 18130 }, { "epoch": 2.306449561124539, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.832029342651367, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8641397953033447, "num_tokens": 691854585.0, "step": 18131 }, { "epoch": 2.3065767714031296, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.765191078186035, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8553320169448853, "num_tokens": 691891784.0, "step": 18132 }, { "epoch": 2.3067039816817196, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.861494064331055, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8745629787445068, "num_tokens": 691927389.0, "step": 18133 }, { "epoch": 2.3068311919603106, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.808297157287598, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8701831102371216, "num_tokens": 691962928.0, "step": 18134 }, { "epoch": 2.3069584022389007, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.790939331054688, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8725587725639343, "num_tokens": 692007064.0, "step": 18135 }, { "epoch": 2.3070856125174912, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.81733512878418, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8763387799263, "num_tokens": 692040668.0, "step": 18136 }, { "epoch": 2.3072128227960818, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.81054401397705, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8737701177597046, "num_tokens": 692077710.0, "step": 18137 }, { "epoch": 2.3073400330746723, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.850602149963379, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.879657506942749, "num_tokens": 692112316.0, "step": 18138 }, { "epoch": 2.307467243353263, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.811532974243164, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8780744075775146, "num_tokens": 692153201.0, "step": 18139 }, { "epoch": 2.3075944536318533, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.821651458740234, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8665950894355774, "num_tokens": 692192942.0, "step": 18140 }, { "epoch": 2.307721663910444, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.845293998718262, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8860553503036499, "num_tokens": 692235515.0, "step": 18141 }, { "epoch": 2.3078488741890344, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.817541122436523, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8639002442359924, "num_tokens": 692271338.0, "step": 18142 }, { "epoch": 2.307976084467625, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.78896427154541, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8534669280052185, "num_tokens": 692311469.0, "step": 18143 }, { "epoch": 2.3081032947462155, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.79493522644043, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8558095097541809, "num_tokens": 692352052.0, "step": 18144 }, { "epoch": 2.308230505024806, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.82903003692627, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8569379448890686, "num_tokens": 692382978.0, "step": 18145 }, { "epoch": 2.3083577153033965, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.798903465270996, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8744347095489502, "num_tokens": 692415651.0, "step": 18146 }, { "epoch": 2.308484925581987, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83644962310791, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8548175096511841, "num_tokens": 692455937.0, "step": 18147 }, { "epoch": 2.3086121358605776, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.907364845275879, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.866492748260498, "num_tokens": 692496277.0, "step": 18148 }, { "epoch": 2.308739346139168, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.818403244018555, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.872838020324707, "num_tokens": 692530485.0, "step": 18149 }, { "epoch": 2.3088665564177586, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.824823379516602, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8692388534545898, "num_tokens": 692564794.0, "step": 18150 }, { "epoch": 2.308993766696349, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.713258743286133, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8620836734771729, "num_tokens": 692610359.0, "step": 18151 }, { "epoch": 2.3091209769749397, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.884666442871094, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8787986636161804, "num_tokens": 692641560.0, "step": 18152 }, { "epoch": 2.30924818725353, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.864026069641113, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8683699369430542, "num_tokens": 692680114.0, "step": 18153 }, { "epoch": 2.3093753975321207, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.798675537109375, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8607608079910278, "num_tokens": 692714199.0, "step": 18154 }, { "epoch": 2.3095026078107113, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.866323471069336, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8741462230682373, "num_tokens": 692752331.0, "step": 18155 }, { "epoch": 2.309629818089302, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.801279067993164, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.878204345703125, "num_tokens": 692789167.0, "step": 18156 }, { "epoch": 2.3097570283678923, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.798059463500977, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8720420598983765, "num_tokens": 692829234.0, "step": 18157 }, { "epoch": 2.3098842386464824, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.870244026184082, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.851912260055542, "num_tokens": 692863749.0, "step": 18158 }, { "epoch": 2.3100114489250734, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.860295295715332, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8675061464309692, "num_tokens": 692900972.0, "step": 18159 }, { "epoch": 2.3101386592036635, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.819628715515137, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8667274713516235, "num_tokens": 692938522.0, "step": 18160 }, { "epoch": 2.310265869482254, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.83598518371582, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8814297914505005, "num_tokens": 692973653.0, "step": 18161 }, { "epoch": 2.3103930797608445, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.826830863952637, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8650072813034058, "num_tokens": 693016325.0, "step": 18162 }, { "epoch": 2.310520290039435, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.879069328308105, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8826242685317993, "num_tokens": 693052022.0, "step": 18163 }, { "epoch": 2.3106475003180256, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.832362174987793, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8768725991249084, "num_tokens": 693093936.0, "step": 18164 }, { "epoch": 2.310774710596616, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.80551815032959, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8853046894073486, "num_tokens": 693130160.0, "step": 18165 }, { "epoch": 2.3109019208752066, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.822526931762695, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.875016450881958, "num_tokens": 693172405.0, "step": 18166 }, { "epoch": 2.311029131153797, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.782816886901855, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8753176927566528, "num_tokens": 693206499.0, "step": 18167 }, { "epoch": 2.3111563414323877, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.856672286987305, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8706197738647461, "num_tokens": 693244287.0, "step": 18168 }, { "epoch": 2.311283551710978, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.85110092163086, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8635214567184448, "num_tokens": 693286164.0, "step": 18169 }, { "epoch": 2.3114107619895687, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.782994270324707, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8786552548408508, "num_tokens": 693330961.0, "step": 18170 }, { "epoch": 2.3115379722681593, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.850674629211426, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8776435852050781, "num_tokens": 693375169.0, "step": 18171 }, { "epoch": 2.31166518254675, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.782649040222168, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.866827130317688, "num_tokens": 693414127.0, "step": 18172 }, { "epoch": 2.3117923928253403, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.819416046142578, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8691225051879883, "num_tokens": 693450446.0, "step": 18173 }, { "epoch": 2.311919603103931, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.779033660888672, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8637838363647461, "num_tokens": 693490973.0, "step": 18174 }, { "epoch": 2.3120468133825214, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.83767032623291, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8745925426483154, "num_tokens": 693526815.0, "step": 18175 }, { "epoch": 2.312174023661112, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.73470687866211, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8660305738449097, "num_tokens": 693566949.0, "step": 18176 }, { "epoch": 2.3123012339397024, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.847566604614258, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.866598904132843, "num_tokens": 693601116.0, "step": 18177 }, { "epoch": 2.312428444218293, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.78276538848877, "learning_rate": 1e-06, "loss": 0.3247, "mean_token_accuracy": 0.9020000696182251, "num_tokens": 693640588.0, "step": 18178 }, { "epoch": 2.3125556544968835, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.781820297241211, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8732324242591858, "num_tokens": 693683352.0, "step": 18179 }, { "epoch": 2.312682864775474, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.856810569763184, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8775055408477783, "num_tokens": 693722579.0, "step": 18180 }, { "epoch": 2.3128100750540646, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.786992073059082, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8734374642372131, "num_tokens": 693766743.0, "step": 18181 }, { "epoch": 2.312937285332655, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.805974960327148, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8750869035720825, "num_tokens": 693809699.0, "step": 18182 }, { "epoch": 2.313064495611245, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.764430046081543, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8778271079063416, "num_tokens": 693845508.0, "step": 18183 }, { "epoch": 2.313191705889836, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.870807647705078, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8714274168014526, "num_tokens": 693883763.0, "step": 18184 }, { "epoch": 2.313318916168426, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.782407760620117, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8776838183403015, "num_tokens": 693917093.0, "step": 18185 }, { "epoch": 2.3134461264470167, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.950288772583008, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8737225532531738, "num_tokens": 693949219.0, "step": 18186 }, { "epoch": 2.3135733367256073, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.80079460144043, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8660608530044556, "num_tokens": 693992755.0, "step": 18187 }, { "epoch": 2.313700547004198, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.932393074035645, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8571293354034424, "num_tokens": 694030386.0, "step": 18188 }, { "epoch": 2.3138277572827883, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.837265014648438, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.880281925201416, "num_tokens": 694065019.0, "step": 18189 }, { "epoch": 2.313954967561379, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.875547409057617, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8699793815612793, "num_tokens": 694106645.0, "step": 18190 }, { "epoch": 2.3140821778399694, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.837961196899414, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8771857023239136, "num_tokens": 694149146.0, "step": 18191 }, { "epoch": 2.31420938811856, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.825103759765625, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8570095896720886, "num_tokens": 694193567.0, "step": 18192 }, { "epoch": 2.3143365983971504, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.818681716918945, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8828791975975037, "num_tokens": 694232160.0, "step": 18193 }, { "epoch": 2.314463808675741, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.763562202453613, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8732740879058838, "num_tokens": 694275153.0, "step": 18194 }, { "epoch": 2.3145910189543315, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.914041519165039, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8789288997650146, "num_tokens": 694313722.0, "step": 18195 }, { "epoch": 2.314718229232922, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.813177108764648, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8764545321464539, "num_tokens": 694350544.0, "step": 18196 }, { "epoch": 2.3148454395115126, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.825217247009277, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8739951848983765, "num_tokens": 694386552.0, "step": 18197 }, { "epoch": 2.314972649790103, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.751315116882324, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8689032793045044, "num_tokens": 694434368.0, "step": 18198 }, { "epoch": 2.3150998600686936, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.817359924316406, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8723415732383728, "num_tokens": 694479856.0, "step": 18199 }, { "epoch": 2.315227070347284, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.855003356933594, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8731771111488342, "num_tokens": 694521107.0, "step": 18200 }, { "epoch": 2.3153542806258747, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.801568031311035, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8733465671539307, "num_tokens": 694560886.0, "step": 18201 }, { "epoch": 2.315481490904465, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.843391418457031, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.876446008682251, "num_tokens": 694598404.0, "step": 18202 }, { "epoch": 2.3156087011830557, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.787416458129883, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8679368495941162, "num_tokens": 694637162.0, "step": 18203 }, { "epoch": 2.3157359114616463, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.812870979309082, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8760849237442017, "num_tokens": 694677435.0, "step": 18204 }, { "epoch": 2.315863121740237, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.784722328186035, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8814852237701416, "num_tokens": 694713980.0, "step": 18205 }, { "epoch": 2.3159903320188273, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.83885669708252, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8679430484771729, "num_tokens": 694753884.0, "step": 18206 }, { "epoch": 2.316117542297418, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.794075965881348, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8851122856140137, "num_tokens": 694785739.0, "step": 18207 }, { "epoch": 2.316244752576008, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.86307144165039, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8716650009155273, "num_tokens": 694827629.0, "step": 18208 }, { "epoch": 2.316371962854599, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.84044075012207, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8685811758041382, "num_tokens": 694868241.0, "step": 18209 }, { "epoch": 2.316499173133189, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.903883934020996, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8634287118911743, "num_tokens": 694905298.0, "step": 18210 }, { "epoch": 2.3166263834117795, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.83791446685791, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8648492097854614, "num_tokens": 694938054.0, "step": 18211 }, { "epoch": 2.31675359369037, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.836664199829102, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8690820932388306, "num_tokens": 694978761.0, "step": 18212 }, { "epoch": 2.3168808039689606, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.799324989318848, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8695613741874695, "num_tokens": 695015073.0, "step": 18213 }, { "epoch": 2.317008014247551, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.838977813720703, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8672696352005005, "num_tokens": 695052288.0, "step": 18214 }, { "epoch": 2.3171352245261416, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.853233337402344, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8623033761978149, "num_tokens": 695088976.0, "step": 18215 }, { "epoch": 2.317262434804732, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.861528396606445, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8752713203430176, "num_tokens": 695128482.0, "step": 18216 }, { "epoch": 2.3173896450833227, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.872025489807129, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8693249225616455, "num_tokens": 695164963.0, "step": 18217 }, { "epoch": 2.317516855361913, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.87246322631836, "learning_rate": 1e-06, "loss": 0.3423, "mean_token_accuracy": 0.8936684131622314, "num_tokens": 695201515.0, "step": 18218 }, { "epoch": 2.3176440656405037, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.88707160949707, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8778955340385437, "num_tokens": 695242057.0, "step": 18219 }, { "epoch": 2.3177712759190943, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83555793762207, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8742982149124146, "num_tokens": 695277128.0, "step": 18220 }, { "epoch": 2.317898486197685, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.792264938354492, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8857196569442749, "num_tokens": 695311984.0, "step": 18221 }, { "epoch": 2.3180256964762753, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.902557373046875, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8622076511383057, "num_tokens": 695348457.0, "step": 18222 }, { "epoch": 2.318152906754866, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.78999137878418, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.874972403049469, "num_tokens": 695386518.0, "step": 18223 }, { "epoch": 2.3182801170334564, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.878151893615723, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8689890503883362, "num_tokens": 695425365.0, "step": 18224 }, { "epoch": 2.318407327312047, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.864518165588379, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8747603893280029, "num_tokens": 695459559.0, "step": 18225 }, { "epoch": 2.3185345375906374, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.839749336242676, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8683528900146484, "num_tokens": 695491386.0, "step": 18226 }, { "epoch": 2.318661747869228, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.84019660949707, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8555670976638794, "num_tokens": 695534660.0, "step": 18227 }, { "epoch": 2.3187889581478185, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.817907333374023, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8645274043083191, "num_tokens": 695578127.0, "step": 18228 }, { "epoch": 2.318916168426409, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.862933158874512, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8605164885520935, "num_tokens": 695614181.0, "step": 18229 }, { "epoch": 2.3190433787049995, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.871841430664062, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8718874454498291, "num_tokens": 695652753.0, "step": 18230 }, { "epoch": 2.3191705889835896, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.908287048339844, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8792747259140015, "num_tokens": 695692449.0, "step": 18231 }, { "epoch": 2.3192977992621806, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.93027400970459, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8715682625770569, "num_tokens": 695730812.0, "step": 18232 }, { "epoch": 2.3194250095407707, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83556842803955, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8872230648994446, "num_tokens": 695767577.0, "step": 18233 }, { "epoch": 2.319552219819361, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.966018676757812, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8723248839378357, "num_tokens": 695812372.0, "step": 18234 }, { "epoch": 2.3196794300979517, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.859123229980469, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.8865463137626648, "num_tokens": 695847470.0, "step": 18235 }, { "epoch": 2.3198066403765423, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.84472370147705, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8878597617149353, "num_tokens": 695888082.0, "step": 18236 }, { "epoch": 2.319933850655133, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.861405372619629, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8801400065422058, "num_tokens": 695928904.0, "step": 18237 }, { "epoch": 2.3200610609337233, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.836715698242188, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8620793223381042, "num_tokens": 695967549.0, "step": 18238 }, { "epoch": 2.320188271212314, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.888778686523438, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8659027814865112, "num_tokens": 696005611.0, "step": 18239 }, { "epoch": 2.3203154814909044, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.926387786865234, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8750736713409424, "num_tokens": 696034864.0, "step": 18240 }, { "epoch": 2.320442691769495, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.916693687438965, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8710631132125854, "num_tokens": 696072137.0, "step": 18241 }, { "epoch": 2.3205699020480854, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.870380401611328, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8752263784408569, "num_tokens": 696115744.0, "step": 18242 }, { "epoch": 2.320697112326676, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.85272216796875, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8686366677284241, "num_tokens": 696151293.0, "step": 18243 }, { "epoch": 2.3208243226052665, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.817662239074707, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8898375034332275, "num_tokens": 696190742.0, "step": 18244 }, { "epoch": 2.320951532883857, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.793149948120117, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8610004186630249, "num_tokens": 696232807.0, "step": 18245 }, { "epoch": 2.3210787431624476, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.886225700378418, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8689426779747009, "num_tokens": 696269416.0, "step": 18246 }, { "epoch": 2.321205953441038, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.786328315734863, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8732734322547913, "num_tokens": 696305144.0, "step": 18247 }, { "epoch": 2.3213331637196286, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.883564949035645, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8643025755882263, "num_tokens": 696344854.0, "step": 18248 }, { "epoch": 2.321460373998219, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.741415977478027, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8910584449768066, "num_tokens": 696380154.0, "step": 18249 }, { "epoch": 2.3215875842768097, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.910998344421387, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8600062131881714, "num_tokens": 696416458.0, "step": 18250 }, { "epoch": 2.3217147945554, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.820261001586914, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8755589127540588, "num_tokens": 696453232.0, "step": 18251 }, { "epoch": 2.3218420048339907, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.865718841552734, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8523257970809937, "num_tokens": 696487930.0, "step": 18252 }, { "epoch": 2.3219692151125813, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.904074668884277, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8637096285820007, "num_tokens": 696523824.0, "step": 18253 }, { "epoch": 2.322096425391172, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.810903549194336, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8819745182991028, "num_tokens": 696567007.0, "step": 18254 }, { "epoch": 2.3222236356697623, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.766775131225586, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8883776068687439, "num_tokens": 696610055.0, "step": 18255 }, { "epoch": 2.3223508459483524, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.737907409667969, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8581193089485168, "num_tokens": 696644058.0, "step": 18256 }, { "epoch": 2.3224780562269434, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.91112995147705, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8797265887260437, "num_tokens": 696686381.0, "step": 18257 }, { "epoch": 2.3226052665055335, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.755682945251465, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8666614890098572, "num_tokens": 696727567.0, "step": 18258 }, { "epoch": 2.322732476784124, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.81985855102539, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8630907535552979, "num_tokens": 696767266.0, "step": 18259 }, { "epoch": 2.3228596870627145, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.766149520874023, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8854985237121582, "num_tokens": 696806059.0, "step": 18260 }, { "epoch": 2.322986897341305, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.847404479980469, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.870976984500885, "num_tokens": 696844232.0, "step": 18261 }, { "epoch": 2.3231141076198956, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.839476585388184, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8744188547134399, "num_tokens": 696886223.0, "step": 18262 }, { "epoch": 2.323241317898486, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.840456008911133, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8586535453796387, "num_tokens": 696922625.0, "step": 18263 }, { "epoch": 2.3233685281770766, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.84969711303711, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8613191843032837, "num_tokens": 696965160.0, "step": 18264 }, { "epoch": 2.323495738455667, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.8228120803833, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8787472248077393, "num_tokens": 697000854.0, "step": 18265 }, { "epoch": 2.3236229487342577, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.761547088623047, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8863814473152161, "num_tokens": 697041264.0, "step": 18266 }, { "epoch": 2.323750159012848, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.891408920288086, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8698430061340332, "num_tokens": 697080079.0, "step": 18267 }, { "epoch": 2.3238773692914387, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.752334594726562, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8690363168716431, "num_tokens": 697123755.0, "step": 18268 }, { "epoch": 2.3240045795700293, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.81004524230957, "learning_rate": 1e-06, "loss": 0.4936, "mean_token_accuracy": 0.8452022075653076, "num_tokens": 697162012.0, "step": 18269 }, { "epoch": 2.32413178984862, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.801019668579102, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8818061351776123, "num_tokens": 697200188.0, "step": 18270 }, { "epoch": 2.3242590001272103, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.799774169921875, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8750603199005127, "num_tokens": 697240536.0, "step": 18271 }, { "epoch": 2.324386210405801, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.852385520935059, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8839857578277588, "num_tokens": 697279459.0, "step": 18272 }, { "epoch": 2.3245134206843914, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.866536140441895, "learning_rate": 1e-06, "loss": 0.3431, "mean_token_accuracy": 0.8888899683952332, "num_tokens": 697312594.0, "step": 18273 }, { "epoch": 2.324640630962982, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.743183135986328, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8596281409263611, "num_tokens": 697357834.0, "step": 18274 }, { "epoch": 2.3247678412415724, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.860486030578613, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8715643882751465, "num_tokens": 697401522.0, "step": 18275 }, { "epoch": 2.324895051520163, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.800226211547852, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8698299527168274, "num_tokens": 697440415.0, "step": 18276 }, { "epoch": 2.3250222617987535, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.893449783325195, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8567686080932617, "num_tokens": 697477609.0, "step": 18277 }, { "epoch": 2.325149472077344, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.799199104309082, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8687950372695923, "num_tokens": 697515566.0, "step": 18278 }, { "epoch": 2.3252766823559345, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.776429176330566, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8767843842506409, "num_tokens": 697549537.0, "step": 18279 }, { "epoch": 2.325403892634525, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.878517150878906, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8768277168273926, "num_tokens": 697582618.0, "step": 18280 }, { "epoch": 2.325531102913115, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.805472373962402, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8860654830932617, "num_tokens": 697622234.0, "step": 18281 }, { "epoch": 2.325658313191706, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.923657417297363, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.863696813583374, "num_tokens": 697657504.0, "step": 18282 }, { "epoch": 2.325785523470296, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.84006118774414, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8598763346672058, "num_tokens": 697699496.0, "step": 18283 }, { "epoch": 2.3259127337488867, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.795793533325195, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8851585388183594, "num_tokens": 697739969.0, "step": 18284 }, { "epoch": 2.3260399440274773, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.94827938079834, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8633360862731934, "num_tokens": 697774346.0, "step": 18285 }, { "epoch": 2.326167154306068, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.807297706604004, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.858835756778717, "num_tokens": 697813013.0, "step": 18286 }, { "epoch": 2.3262943645846583, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.840018272399902, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8661474585533142, "num_tokens": 697854163.0, "step": 18287 }, { "epoch": 2.326421574863249, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.780942916870117, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8796871900558472, "num_tokens": 697898049.0, "step": 18288 }, { "epoch": 2.3265487851418394, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.943875312805176, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8757848143577576, "num_tokens": 697936086.0, "step": 18289 }, { "epoch": 2.32667599542043, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.856632232666016, "learning_rate": 1e-06, "loss": 0.4945, "mean_token_accuracy": 0.8493018746376038, "num_tokens": 697981529.0, "step": 18290 }, { "epoch": 2.3268032056990204, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.851842880249023, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.888966977596283, "num_tokens": 698016047.0, "step": 18291 }, { "epoch": 2.326930415977611, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.753952026367188, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8838106393814087, "num_tokens": 698052408.0, "step": 18292 }, { "epoch": 2.3270576262562015, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.878207206726074, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8653981685638428, "num_tokens": 698089751.0, "step": 18293 }, { "epoch": 2.327184836534792, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.813165664672852, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8707674145698547, "num_tokens": 698121274.0, "step": 18294 }, { "epoch": 2.3273120468133826, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.875528335571289, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8814877271652222, "num_tokens": 698155880.0, "step": 18295 }, { "epoch": 2.327439257091973, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.763821601867676, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8747079968452454, "num_tokens": 698191448.0, "step": 18296 }, { "epoch": 2.3275664673705636, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.90632152557373, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8673597574234009, "num_tokens": 698228437.0, "step": 18297 }, { "epoch": 2.327693677649154, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.769781112670898, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8659192323684692, "num_tokens": 698268714.0, "step": 18298 }, { "epoch": 2.3278208879277447, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.831546783447266, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8729648590087891, "num_tokens": 698305575.0, "step": 18299 }, { "epoch": 2.327948098206335, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.849442481994629, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8675638437271118, "num_tokens": 698347701.0, "step": 18300 }, { "epoch": 2.3280753084849257, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.915979385375977, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8874549865722656, "num_tokens": 698385906.0, "step": 18301 }, { "epoch": 2.3282025187635162, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.874408721923828, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8719182014465332, "num_tokens": 698415511.0, "step": 18302 }, { "epoch": 2.3283297290421068, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.865781784057617, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8677833676338196, "num_tokens": 698460528.0, "step": 18303 }, { "epoch": 2.3284569393206973, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.904647827148438, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8584579825401306, "num_tokens": 698500800.0, "step": 18304 }, { "epoch": 2.328584149599288, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.880675315856934, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8719262480735779, "num_tokens": 698539373.0, "step": 18305 }, { "epoch": 2.328711359877878, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.835742950439453, "learning_rate": 1e-06, "loss": 0.3474, "mean_token_accuracy": 0.8946965932846069, "num_tokens": 698577067.0, "step": 18306 }, { "epoch": 2.328838570156469, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.856485366821289, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8719523549079895, "num_tokens": 698613515.0, "step": 18307 }, { "epoch": 2.328965780435059, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.942012786865234, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8739368915557861, "num_tokens": 698647278.0, "step": 18308 }, { "epoch": 2.3290929907136495, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.859495162963867, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8797198534011841, "num_tokens": 698688738.0, "step": 18309 }, { "epoch": 2.32922020099224, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.905506134033203, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8766809701919556, "num_tokens": 698725003.0, "step": 18310 }, { "epoch": 2.3293474112708306, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.995793342590332, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8544076681137085, "num_tokens": 698768205.0, "step": 18311 }, { "epoch": 2.329474621549421, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.992584228515625, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8530851602554321, "num_tokens": 698805352.0, "step": 18312 }, { "epoch": 2.3296018318280116, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.884492874145508, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8824988603591919, "num_tokens": 698836610.0, "step": 18313 }, { "epoch": 2.329729042106602, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.93511962890625, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8819152116775513, "num_tokens": 698877514.0, "step": 18314 }, { "epoch": 2.3298562523851927, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.914192199707031, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8758725523948669, "num_tokens": 698916699.0, "step": 18315 }, { "epoch": 2.329983462663783, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.84753704071045, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8731803894042969, "num_tokens": 698953568.0, "step": 18316 }, { "epoch": 2.3301106729423737, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.932730674743652, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8676596283912659, "num_tokens": 698990562.0, "step": 18317 }, { "epoch": 2.3302378832209643, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.805658340454102, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8748821020126343, "num_tokens": 699024095.0, "step": 18318 }, { "epoch": 2.330365093499555, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.931061744689941, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.853140652179718, "num_tokens": 699062327.0, "step": 18319 }, { "epoch": 2.3304923037781453, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.823456764221191, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8772021532058716, "num_tokens": 699101591.0, "step": 18320 }, { "epoch": 2.330619514056736, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.937801361083984, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8643170595169067, "num_tokens": 699134836.0, "step": 18321 }, { "epoch": 2.3307467243353264, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.889599800109863, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8628494739532471, "num_tokens": 699171218.0, "step": 18322 }, { "epoch": 2.330873934613917, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.903782844543457, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8901292681694031, "num_tokens": 699200891.0, "step": 18323 }, { "epoch": 2.3310011448925074, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.831324577331543, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8874825239181519, "num_tokens": 699238080.0, "step": 18324 }, { "epoch": 2.331128355171098, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.87479305267334, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8607168793678284, "num_tokens": 699276051.0, "step": 18325 }, { "epoch": 2.3312555654496885, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.837830543518066, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8792776465415955, "num_tokens": 699315721.0, "step": 18326 }, { "epoch": 2.331382775728279, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.846728324890137, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.885006308555603, "num_tokens": 699350342.0, "step": 18327 }, { "epoch": 2.3315099860068695, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.857802391052246, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8767412900924683, "num_tokens": 699390301.0, "step": 18328 }, { "epoch": 2.3316371962854596, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.899112701416016, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8578632473945618, "num_tokens": 699427135.0, "step": 18329 }, { "epoch": 2.3317644065640506, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.888420104980469, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8777925372123718, "num_tokens": 699466461.0, "step": 18330 }, { "epoch": 2.3318916168426407, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.875343322753906, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8657974004745483, "num_tokens": 699502446.0, "step": 18331 }, { "epoch": 2.332018827121231, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.879556655883789, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8723446130752563, "num_tokens": 699546357.0, "step": 18332 }, { "epoch": 2.3321460373998217, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.927688598632812, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8718809485435486, "num_tokens": 699581879.0, "step": 18333 }, { "epoch": 2.3322732476784123, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.779074668884277, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8864198923110962, "num_tokens": 699616287.0, "step": 18334 }, { "epoch": 2.332400457957003, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.822318077087402, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8710068464279175, "num_tokens": 699655409.0, "step": 18335 }, { "epoch": 2.3325276682355933, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.864372253417969, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8473485112190247, "num_tokens": 699701420.0, "step": 18336 }, { "epoch": 2.332654878514184, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.79105281829834, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8743278980255127, "num_tokens": 699736151.0, "step": 18337 }, { "epoch": 2.3327820887927744, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.860010147094727, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8644404411315918, "num_tokens": 699770111.0, "step": 18338 }, { "epoch": 2.332909299071365, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.900656700134277, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8691653609275818, "num_tokens": 699807442.0, "step": 18339 }, { "epoch": 2.3330365093499554, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.762654304504395, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8697326183319092, "num_tokens": 699849364.0, "step": 18340 }, { "epoch": 2.333163719628546, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.871535301208496, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8870676755905151, "num_tokens": 699886234.0, "step": 18341 }, { "epoch": 2.3332909299071365, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.921457290649414, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8552893996238708, "num_tokens": 699925660.0, "step": 18342 }, { "epoch": 2.333418140185727, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.878304481506348, "learning_rate": 1e-06, "loss": 0.4854, "mean_token_accuracy": 0.8458662033081055, "num_tokens": 699967894.0, "step": 18343 }, { "epoch": 2.3335453504643175, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.888710021972656, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8669654130935669, "num_tokens": 699999211.0, "step": 18344 }, { "epoch": 2.333672560742908, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.79055118560791, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8802614212036133, "num_tokens": 700037639.0, "step": 18345 }, { "epoch": 2.3337997710214986, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.91491413116455, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8653242588043213, "num_tokens": 700075677.0, "step": 18346 }, { "epoch": 2.333926981300089, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.791648864746094, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8788678050041199, "num_tokens": 700115697.0, "step": 18347 }, { "epoch": 2.3340541915786797, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.836398124694824, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8552370071411133, "num_tokens": 700152553.0, "step": 18348 }, { "epoch": 2.33418140185727, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.894896507263184, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8794634342193604, "num_tokens": 700190958.0, "step": 18349 }, { "epoch": 2.3343086121358607, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.867521286010742, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8707642555236816, "num_tokens": 700224244.0, "step": 18350 }, { "epoch": 2.3344358224144512, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.841825485229492, "learning_rate": 1e-06, "loss": 0.3234, "mean_token_accuracy": 0.8972874879837036, "num_tokens": 700259736.0, "step": 18351 }, { "epoch": 2.3345630326930418, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.887689590454102, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8726156949996948, "num_tokens": 700293624.0, "step": 18352 }, { "epoch": 2.3346902429716323, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.842689514160156, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8763595223426819, "num_tokens": 700328052.0, "step": 18353 }, { "epoch": 2.3348174532502224, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.907475471496582, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.865510880947113, "num_tokens": 700364652.0, "step": 18354 }, { "epoch": 2.3349446635288134, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.885506629943848, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.88382887840271, "num_tokens": 700403698.0, "step": 18355 }, { "epoch": 2.3350718738074034, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.85472297668457, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8859977722167969, "num_tokens": 700439391.0, "step": 18356 }, { "epoch": 2.335199084085994, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.925646781921387, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.876076340675354, "num_tokens": 700475095.0, "step": 18357 }, { "epoch": 2.3353262943645845, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.821914672851562, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8765670657157898, "num_tokens": 700514167.0, "step": 18358 }, { "epoch": 2.335453504643175, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.848419189453125, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8810079097747803, "num_tokens": 700548333.0, "step": 18359 }, { "epoch": 2.3355807149217656, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.85236644744873, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8772673606872559, "num_tokens": 700594261.0, "step": 18360 }, { "epoch": 2.335707925200356, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.829341888427734, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8781555891036987, "num_tokens": 700632513.0, "step": 18361 }, { "epoch": 2.3358351354789466, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.828591346740723, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8789503574371338, "num_tokens": 700672430.0, "step": 18362 }, { "epoch": 2.335962345757537, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.923089027404785, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8786285519599915, "num_tokens": 700714321.0, "step": 18363 }, { "epoch": 2.3360895560361277, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.8873929977417, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8727961778640747, "num_tokens": 700751714.0, "step": 18364 }, { "epoch": 2.336216766314718, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.866703987121582, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.86972975730896, "num_tokens": 700794087.0, "step": 18365 }, { "epoch": 2.3363439765933087, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.940681457519531, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.867815375328064, "num_tokens": 700832031.0, "step": 18366 }, { "epoch": 2.3364711868718993, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.860465049743652, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8753588795661926, "num_tokens": 700867175.0, "step": 18367 }, { "epoch": 2.33659839715049, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.867198944091797, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8785244226455688, "num_tokens": 700906762.0, "step": 18368 }, { "epoch": 2.3367256074290803, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.867130279541016, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8747376203536987, "num_tokens": 700944316.0, "step": 18369 }, { "epoch": 2.336852817707671, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.876492500305176, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8689143657684326, "num_tokens": 700979369.0, "step": 18370 }, { "epoch": 2.3369800279862614, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.800195693969727, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8694183826446533, "num_tokens": 701016213.0, "step": 18371 }, { "epoch": 2.337107238264852, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.869309425354004, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8712565302848816, "num_tokens": 701055270.0, "step": 18372 }, { "epoch": 2.3372344485434424, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.838874816894531, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8707481622695923, "num_tokens": 701091058.0, "step": 18373 }, { "epoch": 2.337361658822033, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.895331382751465, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8808512091636658, "num_tokens": 701129446.0, "step": 18374 }, { "epoch": 2.3374888691006235, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.80445671081543, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8771228790283203, "num_tokens": 701168726.0, "step": 18375 }, { "epoch": 2.337616079379214, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.852307319641113, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8674032688140869, "num_tokens": 701204041.0, "step": 18376 }, { "epoch": 2.3377432896578045, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.803171157836914, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8511798977851868, "num_tokens": 701244660.0, "step": 18377 }, { "epoch": 2.337870499936395, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.80366325378418, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8639026284217834, "num_tokens": 701284830.0, "step": 18378 }, { "epoch": 2.337997710214985, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.836284637451172, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8800956606864929, "num_tokens": 701319720.0, "step": 18379 }, { "epoch": 2.338124920493576, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.804752349853516, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8516204357147217, "num_tokens": 701361770.0, "step": 18380 }, { "epoch": 2.338252130772166, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.875986099243164, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8700082302093506, "num_tokens": 701399777.0, "step": 18381 }, { "epoch": 2.3383793410507567, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.84083366394043, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8720331192016602, "num_tokens": 701432680.0, "step": 18382 }, { "epoch": 2.3385065513293473, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.84878158569336, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8834848403930664, "num_tokens": 701474101.0, "step": 18383 }, { "epoch": 2.338633761607938, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.86164379119873, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8850146532058716, "num_tokens": 701517568.0, "step": 18384 }, { "epoch": 2.3387609718865283, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.852660179138184, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8665467500686646, "num_tokens": 701558466.0, "step": 18385 }, { "epoch": 2.338888182165119, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.854022979736328, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8682341575622559, "num_tokens": 701591104.0, "step": 18386 }, { "epoch": 2.3390153924437094, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.77262020111084, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.865903377532959, "num_tokens": 701633242.0, "step": 18387 }, { "epoch": 2.3391426027223, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.828761100769043, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8716158866882324, "num_tokens": 701668864.0, "step": 18388 }, { "epoch": 2.3392698130008904, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.79251766204834, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8768419623374939, "num_tokens": 701703648.0, "step": 18389 }, { "epoch": 2.339397023279481, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.857651710510254, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8692022562026978, "num_tokens": 701744417.0, "step": 18390 }, { "epoch": 2.3395242335580715, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.839558601379395, "learning_rate": 1e-06, "loss": 0.4872, "mean_token_accuracy": 0.8409282565116882, "num_tokens": 701783541.0, "step": 18391 }, { "epoch": 2.339651443836662, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.831835746765137, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8683069348335266, "num_tokens": 701825054.0, "step": 18392 }, { "epoch": 2.3397786541152525, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.83851432800293, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8788576126098633, "num_tokens": 701866173.0, "step": 18393 }, { "epoch": 2.339905864393843, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.86491870880127, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8685692548751831, "num_tokens": 701904812.0, "step": 18394 }, { "epoch": 2.3400330746724336, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.824922561645508, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8620832562446594, "num_tokens": 701942590.0, "step": 18395 }, { "epoch": 2.340160284951024, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.926263809204102, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8752121925354004, "num_tokens": 701980773.0, "step": 18396 }, { "epoch": 2.3402874952296147, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.806391716003418, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8857544660568237, "num_tokens": 702014508.0, "step": 18397 }, { "epoch": 2.340414705508205, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.794533729553223, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8847113847732544, "num_tokens": 702055754.0, "step": 18398 }, { "epoch": 2.3405419157867957, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.838930130004883, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.870331883430481, "num_tokens": 702100022.0, "step": 18399 }, { "epoch": 2.3406691260653862, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.809350967407227, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8620466589927673, "num_tokens": 702133217.0, "step": 18400 }, { "epoch": 2.3407963363439768, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.850454330444336, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.884980320930481, "num_tokens": 702176998.0, "step": 18401 }, { "epoch": 2.3409235466225673, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.824549674987793, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8787105679512024, "num_tokens": 702216219.0, "step": 18402 }, { "epoch": 2.341050756901158, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83709716796875, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8674806356430054, "num_tokens": 702246502.0, "step": 18403 }, { "epoch": 2.341177967179748, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.782438278198242, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8794248700141907, "num_tokens": 702277990.0, "step": 18404 }, { "epoch": 2.341305177458339, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.83272933959961, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8749723434448242, "num_tokens": 702312742.0, "step": 18405 }, { "epoch": 2.341432387736929, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.794517517089844, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8874269723892212, "num_tokens": 702348575.0, "step": 18406 }, { "epoch": 2.3415595980155195, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.79985523223877, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8608852624893188, "num_tokens": 702385956.0, "step": 18407 }, { "epoch": 2.34168680829411, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.778388023376465, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8784431219100952, "num_tokens": 702423973.0, "step": 18408 }, { "epoch": 2.3418140185727006, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.799158096313477, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8698106408119202, "num_tokens": 702455508.0, "step": 18409 }, { "epoch": 2.341941228851291, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.909122467041016, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8826124668121338, "num_tokens": 702493365.0, "step": 18410 }, { "epoch": 2.3420684391298816, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.843350410461426, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8747147917747498, "num_tokens": 702528036.0, "step": 18411 }, { "epoch": 2.342195649408472, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.913555145263672, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8677594065666199, "num_tokens": 702571238.0, "step": 18412 }, { "epoch": 2.3423228596870627, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.851431846618652, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8684426546096802, "num_tokens": 702611247.0, "step": 18413 }, { "epoch": 2.342450069965653, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.817743301391602, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.873212456703186, "num_tokens": 702649661.0, "step": 18414 }, { "epoch": 2.3425772802442437, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.814312934875488, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8653950691223145, "num_tokens": 702693573.0, "step": 18415 }, { "epoch": 2.3427044905228342, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.861305236816406, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8840906620025635, "num_tokens": 702735351.0, "step": 18416 }, { "epoch": 2.3428317008014248, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.853848457336426, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8787112236022949, "num_tokens": 702773406.0, "step": 18417 }, { "epoch": 2.3429589110800153, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.81862735748291, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8790609836578369, "num_tokens": 702809939.0, "step": 18418 }, { "epoch": 2.343086121358606, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.866185188293457, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8697477579116821, "num_tokens": 702851120.0, "step": 18419 }, { "epoch": 2.3432133316371964, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.832348823547363, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8809529542922974, "num_tokens": 702889492.0, "step": 18420 }, { "epoch": 2.343340541915787, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.89492416381836, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.867100715637207, "num_tokens": 702930486.0, "step": 18421 }, { "epoch": 2.3434677521943774, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83225154876709, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8646670579910278, "num_tokens": 702969866.0, "step": 18422 }, { "epoch": 2.343594962472968, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.821842193603516, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8643962144851685, "num_tokens": 703013703.0, "step": 18423 }, { "epoch": 2.3437221727515585, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.843873977661133, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8623713254928589, "num_tokens": 703054849.0, "step": 18424 }, { "epoch": 2.343849383030149, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.783839225769043, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8671103715896606, "num_tokens": 703088250.0, "step": 18425 }, { "epoch": 2.3439765933087395, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.833806991577148, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8685969114303589, "num_tokens": 703124072.0, "step": 18426 }, { "epoch": 2.3441038035873296, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.78894329071045, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8555054068565369, "num_tokens": 703158869.0, "step": 18427 }, { "epoch": 2.3442310138659206, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.774953842163086, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8751668930053711, "num_tokens": 703193342.0, "step": 18428 }, { "epoch": 2.3443582241445107, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.830044746398926, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8846358060836792, "num_tokens": 703229899.0, "step": 18429 }, { "epoch": 2.344485434423101, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.848542213439941, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8714334964752197, "num_tokens": 703266735.0, "step": 18430 }, { "epoch": 2.3446126447016917, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.828337669372559, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.868610143661499, "num_tokens": 703307515.0, "step": 18431 }, { "epoch": 2.3447398549802823, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.833184242248535, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8826867341995239, "num_tokens": 703341237.0, "step": 18432 }, { "epoch": 2.344867065258873, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.854159355163574, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8829023838043213, "num_tokens": 703378900.0, "step": 18433 }, { "epoch": 2.3449942755374633, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.823623657226562, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8805734515190125, "num_tokens": 703415260.0, "step": 18434 }, { "epoch": 2.345121485816054, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.883397102355957, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8777871131896973, "num_tokens": 703452399.0, "step": 18435 }, { "epoch": 2.3452486960946444, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.858743667602539, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8739678859710693, "num_tokens": 703492700.0, "step": 18436 }, { "epoch": 2.345375906373235, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.919822692871094, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8841414451599121, "num_tokens": 703530805.0, "step": 18437 }, { "epoch": 2.3455031166518254, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.88424301147461, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8822438716888428, "num_tokens": 703571894.0, "step": 18438 }, { "epoch": 2.345630326930416, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.83709716796875, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8784012794494629, "num_tokens": 703612583.0, "step": 18439 }, { "epoch": 2.3457575372090065, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.861943244934082, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.866305947303772, "num_tokens": 703649100.0, "step": 18440 }, { "epoch": 2.345884747487597, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.911299705505371, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8879718780517578, "num_tokens": 703680425.0, "step": 18441 }, { "epoch": 2.3460119577661875, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.797445297241211, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8789281249046326, "num_tokens": 703721819.0, "step": 18442 }, { "epoch": 2.346139168044778, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.859201431274414, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8696274757385254, "num_tokens": 703755432.0, "step": 18443 }, { "epoch": 2.3462663783233686, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.875640869140625, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.87217777967453, "num_tokens": 703795903.0, "step": 18444 }, { "epoch": 2.346393588601959, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.834574699401855, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8845874071121216, "num_tokens": 703836074.0, "step": 18445 }, { "epoch": 2.3465207988805497, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.843480110168457, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8771722912788391, "num_tokens": 703870609.0, "step": 18446 }, { "epoch": 2.34664800915914, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.886221885681152, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8779036402702332, "num_tokens": 703909034.0, "step": 18447 }, { "epoch": 2.3467752194377307, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.824212074279785, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.865586519241333, "num_tokens": 703946976.0, "step": 18448 }, { "epoch": 2.3469024297163212, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.870146751403809, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8753900527954102, "num_tokens": 703986379.0, "step": 18449 }, { "epoch": 2.3470296399949118, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.832504272460938, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.8915655612945557, "num_tokens": 704019461.0, "step": 18450 }, { "epoch": 2.3471568502735023, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.915340423583984, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8598700165748596, "num_tokens": 704053411.0, "step": 18451 }, { "epoch": 2.3472840605520924, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.86662483215332, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8778445720672607, "num_tokens": 704092633.0, "step": 18452 }, { "epoch": 2.3474112708306833, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.9163236618042, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8778073787689209, "num_tokens": 704126955.0, "step": 18453 }, { "epoch": 2.3475384811092734, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.838691711425781, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.890026867389679, "num_tokens": 704162222.0, "step": 18454 }, { "epoch": 2.347665691387864, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.958418846130371, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8734319806098938, "num_tokens": 704198949.0, "step": 18455 }, { "epoch": 2.3477929016664545, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.904091835021973, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8678988218307495, "num_tokens": 704235327.0, "step": 18456 }, { "epoch": 2.347920111945045, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.881503105163574, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.87920743227005, "num_tokens": 704272783.0, "step": 18457 }, { "epoch": 2.3480473222236355, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.936173439025879, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8568710088729858, "num_tokens": 704314396.0, "step": 18458 }, { "epoch": 2.348174532502226, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.795938491821289, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8900085687637329, "num_tokens": 704346903.0, "step": 18459 }, { "epoch": 2.3483017427808166, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.90212631225586, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8810403943061829, "num_tokens": 704380456.0, "step": 18460 }, { "epoch": 2.348428953059407, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.85070514678955, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8852311372756958, "num_tokens": 704416403.0, "step": 18461 }, { "epoch": 2.3485561633379977, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.900022506713867, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8747613430023193, "num_tokens": 704460431.0, "step": 18462 }, { "epoch": 2.348683373616588, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.863489151000977, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8814812898635864, "num_tokens": 704497872.0, "step": 18463 }, { "epoch": 2.3488105838951787, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.886775016784668, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8761016130447388, "num_tokens": 704538180.0, "step": 18464 }, { "epoch": 2.3489377941737692, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.866891860961914, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8856267929077148, "num_tokens": 704578342.0, "step": 18465 }, { "epoch": 2.3490650044523598, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.85084342956543, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8848513960838318, "num_tokens": 704611792.0, "step": 18466 }, { "epoch": 2.3491922147309503, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.944294929504395, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.876341700553894, "num_tokens": 704649898.0, "step": 18467 }, { "epoch": 2.349319425009541, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.873662948608398, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8828080296516418, "num_tokens": 704690632.0, "step": 18468 }, { "epoch": 2.3494466352881314, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.936125755310059, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8642780780792236, "num_tokens": 704722955.0, "step": 18469 }, { "epoch": 2.349573845566722, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.85502815246582, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8732351660728455, "num_tokens": 704762678.0, "step": 18470 }, { "epoch": 2.3497010558453124, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.842733383178711, "learning_rate": 1e-06, "loss": 0.3232, "mean_token_accuracy": 0.8985307216644287, "num_tokens": 704805440.0, "step": 18471 }, { "epoch": 2.349828266123903, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.929147720336914, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8544948697090149, "num_tokens": 704848153.0, "step": 18472 }, { "epoch": 2.3499554764024935, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.800622940063477, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8716443777084351, "num_tokens": 704886739.0, "step": 18473 }, { "epoch": 2.350082686681084, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.835928916931152, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8735320568084717, "num_tokens": 704926145.0, "step": 18474 }, { "epoch": 2.3502098969596745, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.837922096252441, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.858661413192749, "num_tokens": 704964135.0, "step": 18475 }, { "epoch": 2.350337107238265, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.884303092956543, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8619682788848877, "num_tokens": 705002171.0, "step": 18476 }, { "epoch": 2.350464317516855, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.862000465393066, "learning_rate": 1e-06, "loss": 0.3352, "mean_token_accuracy": 0.8931401968002319, "num_tokens": 705040660.0, "step": 18477 }, { "epoch": 2.350591527795446, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.830984115600586, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8805059194564819, "num_tokens": 705073655.0, "step": 18478 }, { "epoch": 2.350718738074036, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.889060020446777, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.870775580406189, "num_tokens": 705115497.0, "step": 18479 }, { "epoch": 2.3508459483526267, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.844857215881348, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.89069664478302, "num_tokens": 705152642.0, "step": 18480 }, { "epoch": 2.3509731586312173, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.852402687072754, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8595415353775024, "num_tokens": 705187282.0, "step": 18481 }, { "epoch": 2.351100368909808, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.816142082214355, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8770822286605835, "num_tokens": 705225635.0, "step": 18482 }, { "epoch": 2.3512275791883983, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.909241676330566, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8830841183662415, "num_tokens": 705260376.0, "step": 18483 }, { "epoch": 2.351354789466989, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.886021614074707, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8730709552764893, "num_tokens": 705298799.0, "step": 18484 }, { "epoch": 2.3514819997455794, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.907928466796875, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8558283448219299, "num_tokens": 705332570.0, "step": 18485 }, { "epoch": 2.35160921002417, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.901395797729492, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.875946044921875, "num_tokens": 705371214.0, "step": 18486 }, { "epoch": 2.3517364203027604, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.850350379943848, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.871120810508728, "num_tokens": 705405051.0, "step": 18487 }, { "epoch": 2.351863630581351, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.886425018310547, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8712078332901001, "num_tokens": 705442460.0, "step": 18488 }, { "epoch": 2.3519908408599415, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.811330795288086, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8731006383895874, "num_tokens": 705482929.0, "step": 18489 }, { "epoch": 2.352118051138532, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.823604583740234, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8693806529045105, "num_tokens": 705525313.0, "step": 18490 }, { "epoch": 2.3522452614171225, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.842916488647461, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8685408234596252, "num_tokens": 705563730.0, "step": 18491 }, { "epoch": 2.352372471695713, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.872537612915039, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8769205212593079, "num_tokens": 705602459.0, "step": 18492 }, { "epoch": 2.3524996819743036, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.916428565979004, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8715237379074097, "num_tokens": 705636223.0, "step": 18493 }, { "epoch": 2.352626892252894, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.786337852478027, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8725639581680298, "num_tokens": 705676178.0, "step": 18494 }, { "epoch": 2.3527541025314846, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.876633644104004, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8801987767219543, "num_tokens": 705716687.0, "step": 18495 }, { "epoch": 2.352881312810075, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.749242782592773, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8808438777923584, "num_tokens": 705759163.0, "step": 18496 }, { "epoch": 2.3530085230886657, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.94704818725586, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8761496543884277, "num_tokens": 705795982.0, "step": 18497 }, { "epoch": 2.3531357333672562, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.816084861755371, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.877726674079895, "num_tokens": 705835601.0, "step": 18498 }, { "epoch": 2.3532629436458468, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.881677627563477, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8711890578269958, "num_tokens": 705876858.0, "step": 18499 }, { "epoch": 2.353390153924437, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.9175386428833, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8645049929618835, "num_tokens": 705921944.0, "step": 18500 }, { "epoch": 2.353517364203028, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.769522666931152, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8738360404968262, "num_tokens": 705967999.0, "step": 18501 }, { "epoch": 2.353644574481618, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.89656925201416, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8657034635543823, "num_tokens": 706003370.0, "step": 18502 }, { "epoch": 2.353771784760209, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.903544425964355, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8636398911476135, "num_tokens": 706036551.0, "step": 18503 }, { "epoch": 2.353898995038799, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.883872985839844, "learning_rate": 1e-06, "loss": 0.3437, "mean_token_accuracy": 0.891083836555481, "num_tokens": 706075965.0, "step": 18504 }, { "epoch": 2.3540262053173895, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.87163257598877, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8616846799850464, "num_tokens": 706117943.0, "step": 18505 }, { "epoch": 2.35415341559598, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.84256362915039, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8770381212234497, "num_tokens": 706157183.0, "step": 18506 }, { "epoch": 2.3542806258745705, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.848692893981934, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8818057179450989, "num_tokens": 706196878.0, "step": 18507 }, { "epoch": 2.354407836153161, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.90902328491211, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8614215850830078, "num_tokens": 706231102.0, "step": 18508 }, { "epoch": 2.3545350464317516, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.859878540039062, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8816728591918945, "num_tokens": 706273372.0, "step": 18509 }, { "epoch": 2.354662256710342, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.946104049682617, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8670732975006104, "num_tokens": 706308106.0, "step": 18510 }, { "epoch": 2.3547894669889327, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.871210098266602, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8809186816215515, "num_tokens": 706344267.0, "step": 18511 }, { "epoch": 2.354916677267523, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.887572288513184, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8914271593093872, "num_tokens": 706385019.0, "step": 18512 }, { "epoch": 2.3550438875461137, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.897350311279297, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.872471034526825, "num_tokens": 706424913.0, "step": 18513 }, { "epoch": 2.3551710978247042, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.958098411560059, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.873832106590271, "num_tokens": 706466064.0, "step": 18514 }, { "epoch": 2.3552983081032948, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.913253784179688, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8801233768463135, "num_tokens": 706504572.0, "step": 18515 }, { "epoch": 2.3554255183818853, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.901519775390625, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8681454658508301, "num_tokens": 706543901.0, "step": 18516 }, { "epoch": 2.355552728660476, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.96330451965332, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8672629594802856, "num_tokens": 706579095.0, "step": 18517 }, { "epoch": 2.3556799389390664, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.865809440612793, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8671333193778992, "num_tokens": 706614078.0, "step": 18518 }, { "epoch": 2.355807149217657, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.97555923461914, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8809785842895508, "num_tokens": 706652596.0, "step": 18519 }, { "epoch": 2.3559343594962474, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.876370429992676, "learning_rate": 1e-06, "loss": 0.3283, "mean_token_accuracy": 0.8971174359321594, "num_tokens": 706691297.0, "step": 18520 }, { "epoch": 2.356061569774838, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.948460578918457, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8623549938201904, "num_tokens": 706727344.0, "step": 18521 }, { "epoch": 2.3561887800534285, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.858211517333984, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8770034313201904, "num_tokens": 706770411.0, "step": 18522 }, { "epoch": 2.356315990332019, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.851431846618652, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8764034509658813, "num_tokens": 706816215.0, "step": 18523 }, { "epoch": 2.3564432006106095, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.902843475341797, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.868516206741333, "num_tokens": 706857400.0, "step": 18524 }, { "epoch": 2.3565704108891996, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.913675308227539, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8664463758468628, "num_tokens": 706892715.0, "step": 18525 }, { "epoch": 2.3566976211677906, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.925535202026367, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8751794099807739, "num_tokens": 706934373.0, "step": 18526 }, { "epoch": 2.3568248314463807, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.940435409545898, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8749178051948547, "num_tokens": 706976885.0, "step": 18527 }, { "epoch": 2.356952041724971, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.947502136230469, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8796253204345703, "num_tokens": 707016957.0, "step": 18528 }, { "epoch": 2.3570792520035617, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.900403022766113, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8834614753723145, "num_tokens": 707053687.0, "step": 18529 }, { "epoch": 2.3572064622821522, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.88242244720459, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8676754832267761, "num_tokens": 707087327.0, "step": 18530 }, { "epoch": 2.3573336725607428, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.901899337768555, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8852055072784424, "num_tokens": 707124479.0, "step": 18531 }, { "epoch": 2.3574608828393333, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.844683647155762, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8785543441772461, "num_tokens": 707162452.0, "step": 18532 }, { "epoch": 2.357588093117924, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.907588005065918, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8836525082588196, "num_tokens": 707196765.0, "step": 18533 }, { "epoch": 2.3577153033965144, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.830227851867676, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8800607919692993, "num_tokens": 707230151.0, "step": 18534 }, { "epoch": 2.357842513675105, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.879521369934082, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8565053939819336, "num_tokens": 707264373.0, "step": 18535 }, { "epoch": 2.3579697239536954, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.8787202835083, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.86857008934021, "num_tokens": 707304844.0, "step": 18536 }, { "epoch": 2.358096934232286, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.867032051086426, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8760559558868408, "num_tokens": 707347081.0, "step": 18537 }, { "epoch": 2.3582241445108765, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.846861839294434, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8671297430992126, "num_tokens": 707388243.0, "step": 18538 }, { "epoch": 2.358351354789467, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.903379440307617, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8612375259399414, "num_tokens": 707427109.0, "step": 18539 }, { "epoch": 2.3584785650680575, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.881855010986328, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8763021230697632, "num_tokens": 707462502.0, "step": 18540 }, { "epoch": 2.358605775346648, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.864033699035645, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8682809472084045, "num_tokens": 707502861.0, "step": 18541 }, { "epoch": 2.3587329856252386, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.959209442138672, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8535895943641663, "num_tokens": 707535868.0, "step": 18542 }, { "epoch": 2.358860195903829, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.941618919372559, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8770552277565002, "num_tokens": 707566128.0, "step": 18543 }, { "epoch": 2.3589874061824196, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.82125186920166, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8808992505073547, "num_tokens": 707605113.0, "step": 18544 }, { "epoch": 2.35911461646101, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.917856216430664, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8725683689117432, "num_tokens": 707647968.0, "step": 18545 }, { "epoch": 2.3592418267396007, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.929428100585938, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8813273906707764, "num_tokens": 707682183.0, "step": 18546 }, { "epoch": 2.3593690370181912, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.904257774353027, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8816640377044678, "num_tokens": 707720938.0, "step": 18547 }, { "epoch": 2.3594962472967818, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.816289901733398, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8621265888214111, "num_tokens": 707762624.0, "step": 18548 }, { "epoch": 2.3596234575753723, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.959550857543945, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8666287064552307, "num_tokens": 707798406.0, "step": 18549 }, { "epoch": 2.3597506678539624, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.864612579345703, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8759168386459351, "num_tokens": 707835935.0, "step": 18550 }, { "epoch": 2.3598778781325533, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.828558921813965, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.885223388671875, "num_tokens": 707872971.0, "step": 18551 }, { "epoch": 2.3600050884111434, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.921422958374023, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8846509456634521, "num_tokens": 707909798.0, "step": 18552 }, { "epoch": 2.360132298689734, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.838006019592285, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8770157694816589, "num_tokens": 707951130.0, "step": 18553 }, { "epoch": 2.3602595089683245, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.900177955627441, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8619976043701172, "num_tokens": 707989502.0, "step": 18554 }, { "epoch": 2.360386719246915, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.884272575378418, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8730212450027466, "num_tokens": 708023521.0, "step": 18555 }, { "epoch": 2.3605139295255055, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.827646255493164, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8800327777862549, "num_tokens": 708057594.0, "step": 18556 }, { "epoch": 2.360641139804096, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.928793907165527, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8713263273239136, "num_tokens": 708101393.0, "step": 18557 }, { "epoch": 2.3607683500826866, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.831171989440918, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8847084045410156, "num_tokens": 708137141.0, "step": 18558 }, { "epoch": 2.360895560361277, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.919368743896484, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8790649175643921, "num_tokens": 708173497.0, "step": 18559 }, { "epoch": 2.3610227706398677, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.881712913513184, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8727926015853882, "num_tokens": 708209412.0, "step": 18560 }, { "epoch": 2.361149980918458, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.842146873474121, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8819369077682495, "num_tokens": 708239675.0, "step": 18561 }, { "epoch": 2.3612771911970487, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.895047187805176, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.868503987789154, "num_tokens": 708274918.0, "step": 18562 }, { "epoch": 2.3614044014756392, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.899978637695312, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8592195510864258, "num_tokens": 708313645.0, "step": 18563 }, { "epoch": 2.3615316117542298, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.913368225097656, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.874342143535614, "num_tokens": 708357985.0, "step": 18564 }, { "epoch": 2.3616588220328203, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.881601333618164, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8534131050109863, "num_tokens": 708393606.0, "step": 18565 }, { "epoch": 2.361786032311411, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.882867813110352, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8796044588088989, "num_tokens": 708428941.0, "step": 18566 }, { "epoch": 2.3619132425900013, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.874298095703125, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8780417442321777, "num_tokens": 708467847.0, "step": 18567 }, { "epoch": 2.362040452868592, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.886040687561035, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8898205757141113, "num_tokens": 708503212.0, "step": 18568 }, { "epoch": 2.3621676631471824, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.86230182647705, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8742713928222656, "num_tokens": 708542897.0, "step": 18569 }, { "epoch": 2.362294873425773, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.93938159942627, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.863105833530426, "num_tokens": 708574725.0, "step": 18570 }, { "epoch": 2.3624220837043635, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.899131774902344, "learning_rate": 1e-06, "loss": 0.3601, "mean_token_accuracy": 0.8846803307533264, "num_tokens": 708613557.0, "step": 18571 }, { "epoch": 2.362549293982954, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.940589904785156, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8661922216415405, "num_tokens": 708651872.0, "step": 18572 }, { "epoch": 2.3626765042615445, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.91542911529541, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8618566989898682, "num_tokens": 708688654.0, "step": 18573 }, { "epoch": 2.362803714540135, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.934611320495605, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8804060816764832, "num_tokens": 708723285.0, "step": 18574 }, { "epoch": 2.362930924818725, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.917193412780762, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8829272985458374, "num_tokens": 708760617.0, "step": 18575 }, { "epoch": 2.363058135097316, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.947073936462402, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8789693117141724, "num_tokens": 708790792.0, "step": 18576 }, { "epoch": 2.363185345375906, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.920649528503418, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8647856116294861, "num_tokens": 708829935.0, "step": 18577 }, { "epoch": 2.3633125556544967, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.881969451904297, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8692367076873779, "num_tokens": 708870578.0, "step": 18578 }, { "epoch": 2.3634397659330872, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.997210502624512, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8724139332771301, "num_tokens": 708907044.0, "step": 18579 }, { "epoch": 2.3635669762116778, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.928030967712402, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8563094139099121, "num_tokens": 708947443.0, "step": 18580 }, { "epoch": 2.3636941864902683, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.857442855834961, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.875434160232544, "num_tokens": 708983174.0, "step": 18581 }, { "epoch": 2.363821396768859, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.926324844360352, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8767918348312378, "num_tokens": 709021950.0, "step": 18582 }, { "epoch": 2.3639486070474494, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.881208419799805, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8798704147338867, "num_tokens": 709057236.0, "step": 18583 }, { "epoch": 2.36407581732604, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.953347206115723, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8748772144317627, "num_tokens": 709093376.0, "step": 18584 }, { "epoch": 2.3642030276046304, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.902236938476562, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8694573640823364, "num_tokens": 709129502.0, "step": 18585 }, { "epoch": 2.364330237883221, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.015776634216309, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8736876845359802, "num_tokens": 709162538.0, "step": 18586 }, { "epoch": 2.3644574481618115, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.87917709350586, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8634538650512695, "num_tokens": 709201690.0, "step": 18587 }, { "epoch": 2.364584658440402, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.96444320678711, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8673762083053589, "num_tokens": 709240593.0, "step": 18588 }, { "epoch": 2.3647118687189925, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.877644538879395, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8848709464073181, "num_tokens": 709278192.0, "step": 18589 }, { "epoch": 2.364839078997583, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.967952728271484, "learning_rate": 1e-06, "loss": 0.3359, "mean_token_accuracy": 0.8913174867630005, "num_tokens": 709311797.0, "step": 18590 }, { "epoch": 2.3649662892761736, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.802939414978027, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8717021942138672, "num_tokens": 709350379.0, "step": 18591 }, { "epoch": 2.365093499554764, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.936576843261719, "learning_rate": 1e-06, "loss": 0.3223, "mean_token_accuracy": 0.8984485864639282, "num_tokens": 709389030.0, "step": 18592 }, { "epoch": 2.3652207098333546, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.866470336914062, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8659547567367554, "num_tokens": 709433692.0, "step": 18593 }, { "epoch": 2.365347920111945, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.959503173828125, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8739598393440247, "num_tokens": 709470773.0, "step": 18594 }, { "epoch": 2.3654751303905357, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.96810245513916, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8697499632835388, "num_tokens": 709506764.0, "step": 18595 }, { "epoch": 2.3656023406691262, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.918085098266602, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8725837469100952, "num_tokens": 709546931.0, "step": 18596 }, { "epoch": 2.3657295509477168, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.994109153747559, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8754114508628845, "num_tokens": 709582842.0, "step": 18597 }, { "epoch": 2.365856761226307, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.890816688537598, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8734673857688904, "num_tokens": 709621574.0, "step": 18598 }, { "epoch": 2.365983971504898, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.854039192199707, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8916653394699097, "num_tokens": 709662437.0, "step": 18599 }, { "epoch": 2.366111181783488, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.95716667175293, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8872851729393005, "num_tokens": 709701318.0, "step": 18600 }, { "epoch": 2.3662383920620784, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.863229751586914, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8733848333358765, "num_tokens": 709745156.0, "step": 18601 }, { "epoch": 2.366365602340669, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.877456665039062, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8691515922546387, "num_tokens": 709785849.0, "step": 18602 }, { "epoch": 2.3664928126192595, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.909366607666016, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8777934908866882, "num_tokens": 709830038.0, "step": 18603 }, { "epoch": 2.36662002289785, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.821122169494629, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8721336722373962, "num_tokens": 709871174.0, "step": 18604 }, { "epoch": 2.3667472331764405, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.967455863952637, "learning_rate": 1e-06, "loss": 0.3317, "mean_token_accuracy": 0.8948312997817993, "num_tokens": 709905921.0, "step": 18605 }, { "epoch": 2.366874443455031, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.892020225524902, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8898874521255493, "num_tokens": 709939856.0, "step": 18606 }, { "epoch": 2.3670016537336216, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.89773178100586, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8861359357833862, "num_tokens": 709978805.0, "step": 18607 }, { "epoch": 2.367128864012212, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.900260925292969, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8877378106117249, "num_tokens": 710015015.0, "step": 18608 }, { "epoch": 2.3672560742908026, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.928831100463867, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.865465521812439, "num_tokens": 710054887.0, "step": 18609 }, { "epoch": 2.367383284569393, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.97782039642334, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8742238283157349, "num_tokens": 710097012.0, "step": 18610 }, { "epoch": 2.3675104948479837, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.903764724731445, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8583598136901855, "num_tokens": 710130912.0, "step": 18611 }, { "epoch": 2.3676377051265742, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.928025245666504, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8700775504112244, "num_tokens": 710168880.0, "step": 18612 }, { "epoch": 2.3677649154051648, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.97141170501709, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8833339214324951, "num_tokens": 710205619.0, "step": 18613 }, { "epoch": 2.3678921256837553, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.949763298034668, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8798059225082397, "num_tokens": 710239475.0, "step": 18614 }, { "epoch": 2.368019335962346, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.907873153686523, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8765691518783569, "num_tokens": 710276533.0, "step": 18615 }, { "epoch": 2.3681465462409363, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.906229972839355, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.889829695224762, "num_tokens": 710316731.0, "step": 18616 }, { "epoch": 2.368273756519527, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.920963287353516, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8713709115982056, "num_tokens": 710352310.0, "step": 18617 }, { "epoch": 2.3684009667981174, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.908649444580078, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8705325126647949, "num_tokens": 710400203.0, "step": 18618 }, { "epoch": 2.368528177076708, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.892390251159668, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8810461163520813, "num_tokens": 710433054.0, "step": 18619 }, { "epoch": 2.3686553873552985, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.950096130371094, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8781833648681641, "num_tokens": 710466879.0, "step": 18620 }, { "epoch": 2.368782597633889, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.929533958435059, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8681086301803589, "num_tokens": 710502808.0, "step": 18621 }, { "epoch": 2.3689098079124795, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.91305160522461, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8640028238296509, "num_tokens": 710546190.0, "step": 18622 }, { "epoch": 2.3690370181910696, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.918679237365723, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8767601251602173, "num_tokens": 710585556.0, "step": 18623 }, { "epoch": 2.3691642284696606, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.889195442199707, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8773988485336304, "num_tokens": 710621366.0, "step": 18624 }, { "epoch": 2.3692914387482507, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.924971580505371, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8789416551589966, "num_tokens": 710660217.0, "step": 18625 }, { "epoch": 2.369418649026841, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.917482376098633, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.871455192565918, "num_tokens": 710695798.0, "step": 18626 }, { "epoch": 2.3695458593054317, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.862635612487793, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8578121662139893, "num_tokens": 710732739.0, "step": 18627 }, { "epoch": 2.3696730695840222, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.909383773803711, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8670580387115479, "num_tokens": 710769651.0, "step": 18628 }, { "epoch": 2.3698002798626128, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.9500093460083, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8728082776069641, "num_tokens": 710809487.0, "step": 18629 }, { "epoch": 2.3699274901412033, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.019048690795898, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8772426247596741, "num_tokens": 710849152.0, "step": 18630 }, { "epoch": 2.370054700419794, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.873010635375977, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.860992431640625, "num_tokens": 710890933.0, "step": 18631 }, { "epoch": 2.3701819106983844, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.967097282409668, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8775589466094971, "num_tokens": 710928262.0, "step": 18632 }, { "epoch": 2.370309120976975, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.983717918395996, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8794969320297241, "num_tokens": 710962969.0, "step": 18633 }, { "epoch": 2.3704363312555654, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.969612121582031, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8701289892196655, "num_tokens": 710998717.0, "step": 18634 }, { "epoch": 2.370563541534156, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.96298885345459, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8595906496047974, "num_tokens": 711032819.0, "step": 18635 }, { "epoch": 2.3706907518127465, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.992809295654297, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8648405075073242, "num_tokens": 711069395.0, "step": 18636 }, { "epoch": 2.370817962091337, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.916790008544922, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8763206005096436, "num_tokens": 711105754.0, "step": 18637 }, { "epoch": 2.3709451723699275, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.86871337890625, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8688185214996338, "num_tokens": 711145291.0, "step": 18638 }, { "epoch": 2.371072382648518, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.910167694091797, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8757844567298889, "num_tokens": 711183431.0, "step": 18639 }, { "epoch": 2.3711995929271086, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.883031845092773, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8599569797515869, "num_tokens": 711229754.0, "step": 18640 }, { "epoch": 2.371326803205699, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.861945152282715, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8752824664115906, "num_tokens": 711269071.0, "step": 18641 }, { "epoch": 2.3714540134842896, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.98012638092041, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8711929321289062, "num_tokens": 711312036.0, "step": 18642 }, { "epoch": 2.37158122376288, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.870794296264648, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8814870119094849, "num_tokens": 711347945.0, "step": 18643 }, { "epoch": 2.3717084340414707, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.947575569152832, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8531635999679565, "num_tokens": 711388433.0, "step": 18644 }, { "epoch": 2.371835644320061, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.902220726013184, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8785836696624756, "num_tokens": 711425821.0, "step": 18645 }, { "epoch": 2.3719628545986517, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.937994003295898, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8680214881896973, "num_tokens": 711462725.0, "step": 18646 }, { "epoch": 2.3720900648772423, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.94234848022461, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.887953519821167, "num_tokens": 711499987.0, "step": 18647 }, { "epoch": 2.3722172751558324, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.980088233947754, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8593779802322388, "num_tokens": 711543658.0, "step": 18648 }, { "epoch": 2.3723444854344233, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.840177536010742, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.885339617729187, "num_tokens": 711583875.0, "step": 18649 }, { "epoch": 2.3724716957130134, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.90009880065918, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8571820259094238, "num_tokens": 711625543.0, "step": 18650 }, { "epoch": 2.372598905991604, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.967230796813965, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8771227598190308, "num_tokens": 711660528.0, "step": 18651 }, { "epoch": 2.3727261162701945, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.9857816696167, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8898197412490845, "num_tokens": 711695646.0, "step": 18652 }, { "epoch": 2.372853326548785, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.98009967803955, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8591998815536499, "num_tokens": 711734691.0, "step": 18653 }, { "epoch": 2.3729805368273755, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.940827369689941, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8758460283279419, "num_tokens": 711771518.0, "step": 18654 }, { "epoch": 2.373107747105966, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.93305778503418, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8635884523391724, "num_tokens": 711808968.0, "step": 18655 }, { "epoch": 2.3732349573845566, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.99992847442627, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8813287019729614, "num_tokens": 711842944.0, "step": 18656 }, { "epoch": 2.373362167663147, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.894806861877441, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8712611198425293, "num_tokens": 711882740.0, "step": 18657 }, { "epoch": 2.3734893779417376, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.904576301574707, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.867021918296814, "num_tokens": 711925362.0, "step": 18658 }, { "epoch": 2.373616588220328, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.87222957611084, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.877708911895752, "num_tokens": 711971285.0, "step": 18659 }, { "epoch": 2.3737437984989187, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.8869047164917, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8695262670516968, "num_tokens": 712013306.0, "step": 18660 }, { "epoch": 2.3738710087775092, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.947518348693848, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8751950860023499, "num_tokens": 712050837.0, "step": 18661 }, { "epoch": 2.3739982190560998, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.91286563873291, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8904633522033691, "num_tokens": 712085140.0, "step": 18662 }, { "epoch": 2.3741254293346903, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.900341987609863, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8874156475067139, "num_tokens": 712122495.0, "step": 18663 }, { "epoch": 2.374252639613281, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.928573608398438, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8681318759918213, "num_tokens": 712161312.0, "step": 18664 }, { "epoch": 2.3743798498918713, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.899782180786133, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8736296892166138, "num_tokens": 712197275.0, "step": 18665 }, { "epoch": 2.374507060170462, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.881678581237793, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8619471192359924, "num_tokens": 712231309.0, "step": 18666 }, { "epoch": 2.3746342704490524, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.9701566696167, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.869473397731781, "num_tokens": 712264864.0, "step": 18667 }, { "epoch": 2.374761480727643, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.908744812011719, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8736759424209595, "num_tokens": 712303564.0, "step": 18668 }, { "epoch": 2.3748886910062335, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.93921947479248, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8845267295837402, "num_tokens": 712341872.0, "step": 18669 }, { "epoch": 2.375015901284824, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.934223175048828, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8627749681472778, "num_tokens": 712379835.0, "step": 18670 }, { "epoch": 2.3751431115634145, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.838186264038086, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8700955510139465, "num_tokens": 712417462.0, "step": 18671 }, { "epoch": 2.375270321842005, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.976764678955078, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8496096134185791, "num_tokens": 712461581.0, "step": 18672 }, { "epoch": 2.375397532120595, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.873836517333984, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8725059032440186, "num_tokens": 712499615.0, "step": 18673 }, { "epoch": 2.375524742399186, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.902251243591309, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8840720653533936, "num_tokens": 712533264.0, "step": 18674 }, { "epoch": 2.375651952677776, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.871277809143066, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8659266233444214, "num_tokens": 712571496.0, "step": 18675 }, { "epoch": 2.3757791629563667, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.850043296813965, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8803426623344421, "num_tokens": 712606510.0, "step": 18676 }, { "epoch": 2.3759063732349572, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.960613250732422, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8827285766601562, "num_tokens": 712646218.0, "step": 18677 }, { "epoch": 2.3760335835135478, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.823640823364258, "learning_rate": 1e-06, "loss": 0.4965, "mean_token_accuracy": 0.8434046506881714, "num_tokens": 712681905.0, "step": 18678 }, { "epoch": 2.3761607937921383, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.903745651245117, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.8860944509506226, "num_tokens": 712719916.0, "step": 18679 }, { "epoch": 2.376288004070729, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.941399574279785, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8741300106048584, "num_tokens": 712760849.0, "step": 18680 }, { "epoch": 2.3764152143493193, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.91656494140625, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8847845792770386, "num_tokens": 712801199.0, "step": 18681 }, { "epoch": 2.37654242462791, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.932692527770996, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8677011728286743, "num_tokens": 712843471.0, "step": 18682 }, { "epoch": 2.3766696349065004, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.921247482299805, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8742965459823608, "num_tokens": 712880649.0, "step": 18683 }, { "epoch": 2.376796845185091, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.0003023147583, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8506090641021729, "num_tokens": 712920574.0, "step": 18684 }, { "epoch": 2.3769240554636815, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.949559211730957, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8823437094688416, "num_tokens": 712962544.0, "step": 18685 }, { "epoch": 2.377051265742272, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.93497085571289, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8780792355537415, "num_tokens": 712999138.0, "step": 18686 }, { "epoch": 2.3771784760208625, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.833158493041992, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8756327629089355, "num_tokens": 713041799.0, "step": 18687 }, { "epoch": 2.377305686299453, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.964458465576172, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8667733669281006, "num_tokens": 713088248.0, "step": 18688 }, { "epoch": 2.3774328965780436, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.879375457763672, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8832231760025024, "num_tokens": 713127239.0, "step": 18689 }, { "epoch": 2.377560106856634, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.965325355529785, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8770954608917236, "num_tokens": 713168389.0, "step": 18690 }, { "epoch": 2.3776873171352246, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.875361442565918, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8669661283493042, "num_tokens": 713208137.0, "step": 18691 }, { "epoch": 2.377814527413815, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.929865837097168, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8646095991134644, "num_tokens": 713249344.0, "step": 18692 }, { "epoch": 2.3779417376924057, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.941073417663574, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8694817423820496, "num_tokens": 713291376.0, "step": 18693 }, { "epoch": 2.378068947970996, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.901178359985352, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8720740079879761, "num_tokens": 713329925.0, "step": 18694 }, { "epoch": 2.3781961582495867, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.924715995788574, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8762309551239014, "num_tokens": 713372509.0, "step": 18695 }, { "epoch": 2.378323368528177, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.881487846374512, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8679452538490295, "num_tokens": 713410055.0, "step": 18696 }, { "epoch": 2.378450578806768, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.928719520568848, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8718522191047668, "num_tokens": 713453958.0, "step": 18697 }, { "epoch": 2.378577789085358, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.900818824768066, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8740785121917725, "num_tokens": 713485683.0, "step": 18698 }, { "epoch": 2.3787049993639484, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.897242546081543, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8749221563339233, "num_tokens": 713528565.0, "step": 18699 }, { "epoch": 2.378832209642539, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.993640899658203, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8698821663856506, "num_tokens": 713560932.0, "step": 18700 }, { "epoch": 2.3789594199211295, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.875686645507812, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8710411190986633, "num_tokens": 713592891.0, "step": 18701 }, { "epoch": 2.37908663019972, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.923628807067871, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.871062695980072, "num_tokens": 713631285.0, "step": 18702 }, { "epoch": 2.3792138404783105, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.906166076660156, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8561619520187378, "num_tokens": 713675143.0, "step": 18703 }, { "epoch": 2.379341050756901, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.860404014587402, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8775887489318848, "num_tokens": 713717556.0, "step": 18704 }, { "epoch": 2.3794682610354916, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.876521110534668, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8662112951278687, "num_tokens": 713763515.0, "step": 18705 }, { "epoch": 2.379595471314082, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.931868553161621, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8790090084075928, "num_tokens": 713801064.0, "step": 18706 }, { "epoch": 2.3797226815926726, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.94579792022705, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8704588413238525, "num_tokens": 713841662.0, "step": 18707 }, { "epoch": 2.379849891871263, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.89486026763916, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8752480745315552, "num_tokens": 713883646.0, "step": 18708 }, { "epoch": 2.3799771021498537, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.90410041809082, "learning_rate": 1e-06, "loss": 0.3439, "mean_token_accuracy": 0.8933055400848389, "num_tokens": 713925365.0, "step": 18709 }, { "epoch": 2.3801043124284442, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.920221328735352, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8699802756309509, "num_tokens": 713961142.0, "step": 18710 }, { "epoch": 2.3802315227070348, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.968622207641602, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8814444541931152, "num_tokens": 714001652.0, "step": 18711 }, { "epoch": 2.3803587329856253, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.900870323181152, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8752332925796509, "num_tokens": 714043023.0, "step": 18712 }, { "epoch": 2.380485943264216, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.927900314331055, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8728089928627014, "num_tokens": 714080389.0, "step": 18713 }, { "epoch": 2.3806131535428063, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.927054405212402, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8780432939529419, "num_tokens": 714117695.0, "step": 18714 }, { "epoch": 2.380740363821397, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.977569580078125, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8807663321495056, "num_tokens": 714154204.0, "step": 18715 }, { "epoch": 2.3808675740999874, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.93056583404541, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.856665313243866, "num_tokens": 714194857.0, "step": 18716 }, { "epoch": 2.380994784378578, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.977633476257324, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8850511312484741, "num_tokens": 714228034.0, "step": 18717 }, { "epoch": 2.3811219946571685, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.911894798278809, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8803238272666931, "num_tokens": 714269688.0, "step": 18718 }, { "epoch": 2.381249204935759, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 14.025042533874512, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8690661191940308, "num_tokens": 714309039.0, "step": 18719 }, { "epoch": 2.3813764152143495, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.995369911193848, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8789408206939697, "num_tokens": 714347453.0, "step": 18720 }, { "epoch": 2.3815036254929396, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.911052703857422, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8720787763595581, "num_tokens": 714385850.0, "step": 18721 }, { "epoch": 2.3816308357715306, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.996683120727539, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8683105707168579, "num_tokens": 714426735.0, "step": 18722 }, { "epoch": 2.3817580460501206, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.947562217712402, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8848511576652527, "num_tokens": 714465279.0, "step": 18723 }, { "epoch": 2.381885256328711, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.905917167663574, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8631104230880737, "num_tokens": 714505703.0, "step": 18724 }, { "epoch": 2.3820124666073017, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.916831016540527, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8671689629554749, "num_tokens": 714543642.0, "step": 18725 }, { "epoch": 2.3821396768858922, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.93138313293457, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8905467987060547, "num_tokens": 714583533.0, "step": 18726 }, { "epoch": 2.3822668871644828, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.95003890991211, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8787033557891846, "num_tokens": 714625577.0, "step": 18727 }, { "epoch": 2.3823940974430733, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.930603981018066, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8705601096153259, "num_tokens": 714664987.0, "step": 18728 }, { "epoch": 2.382521307721664, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.864350318908691, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8679356575012207, "num_tokens": 714705881.0, "step": 18729 }, { "epoch": 2.3826485180002543, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.011263847351074, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8646979928016663, "num_tokens": 714746012.0, "step": 18730 }, { "epoch": 2.382775728278845, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.88866138458252, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8759214878082275, "num_tokens": 714790001.0, "step": 18731 }, { "epoch": 2.3829029385574354, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.972139358520508, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8802194595336914, "num_tokens": 714824515.0, "step": 18732 }, { "epoch": 2.383030148836026, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.918438911437988, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.866136908531189, "num_tokens": 714860841.0, "step": 18733 }, { "epoch": 2.3831573591146165, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.962735176086426, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8760759830474854, "num_tokens": 714896373.0, "step": 18734 }, { "epoch": 2.383284569393207, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.967655181884766, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.876396119594574, "num_tokens": 714926684.0, "step": 18735 }, { "epoch": 2.3834117796717975, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.865400314331055, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8879109621047974, "num_tokens": 714961057.0, "step": 18736 }, { "epoch": 2.383538989950388, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.997984886169434, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.87906813621521, "num_tokens": 715001215.0, "step": 18737 }, { "epoch": 2.3836662002289786, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.926850318908691, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8546298742294312, "num_tokens": 715038252.0, "step": 18738 }, { "epoch": 2.383793410507569, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.974289894104004, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8735276460647583, "num_tokens": 715078320.0, "step": 18739 }, { "epoch": 2.3839206207861596, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.924553871154785, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8681015372276306, "num_tokens": 715115628.0, "step": 18740 }, { "epoch": 2.38404783106475, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.90766716003418, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8831087350845337, "num_tokens": 715153294.0, "step": 18741 }, { "epoch": 2.3841750413433407, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.909143447875977, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8741089105606079, "num_tokens": 715193736.0, "step": 18742 }, { "epoch": 2.384302251621931, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.943310737609863, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8747657537460327, "num_tokens": 715234539.0, "step": 18743 }, { "epoch": 2.3844294619005217, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.88674259185791, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8863773941993713, "num_tokens": 715271567.0, "step": 18744 }, { "epoch": 2.3845566721791123, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.966769218444824, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8860694766044617, "num_tokens": 715314850.0, "step": 18745 }, { "epoch": 2.3846838824577024, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.91401195526123, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8802623748779297, "num_tokens": 715353353.0, "step": 18746 }, { "epoch": 2.3848110927362933, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 14.023764610290527, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8854743242263794, "num_tokens": 715387843.0, "step": 18747 }, { "epoch": 2.3849383030148834, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.94914722442627, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8616369962692261, "num_tokens": 715428037.0, "step": 18748 }, { "epoch": 2.385065513293474, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.864462852478027, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8881123065948486, "num_tokens": 715464403.0, "step": 18749 }, { "epoch": 2.3851927235720645, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.943136215209961, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8834995031356812, "num_tokens": 715501820.0, "step": 18750 }, { "epoch": 2.385319933850655, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.892903327941895, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8832623958587646, "num_tokens": 715543287.0, "step": 18751 }, { "epoch": 2.3854471441292455, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.001208305358887, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8835314512252808, "num_tokens": 715578189.0, "step": 18752 }, { "epoch": 2.385574354407836, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.975350379943848, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8650122880935669, "num_tokens": 715616447.0, "step": 18753 }, { "epoch": 2.3857015646864266, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.93174934387207, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8784081935882568, "num_tokens": 715651692.0, "step": 18754 }, { "epoch": 2.385828774965017, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.998726844787598, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8674001097679138, "num_tokens": 715687733.0, "step": 18755 }, { "epoch": 2.3859559852436076, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.956411361694336, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8519598245620728, "num_tokens": 715727481.0, "step": 18756 }, { "epoch": 2.386083195522198, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.916499137878418, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.857707142829895, "num_tokens": 715767549.0, "step": 18757 }, { "epoch": 2.3862104058007887, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.942169189453125, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8677324652671814, "num_tokens": 715807251.0, "step": 18758 }, { "epoch": 2.386337616079379, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.9616060256958, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8587143421173096, "num_tokens": 715846893.0, "step": 18759 }, { "epoch": 2.3864648263579697, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.89450454711914, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8613983392715454, "num_tokens": 715883611.0, "step": 18760 }, { "epoch": 2.3865920366365603, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.929101943969727, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8762222528457642, "num_tokens": 715915598.0, "step": 18761 }, { "epoch": 2.386719246915151, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.919950485229492, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8811848163604736, "num_tokens": 715956647.0, "step": 18762 }, { "epoch": 2.3868464571937413, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.99881649017334, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8811274766921997, "num_tokens": 715996627.0, "step": 18763 }, { "epoch": 2.386973667472332, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.985870361328125, "learning_rate": 1e-06, "loss": 0.3313, "mean_token_accuracy": 0.8937903642654419, "num_tokens": 716027823.0, "step": 18764 }, { "epoch": 2.3871008777509224, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.952936172485352, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8608511090278625, "num_tokens": 716067430.0, "step": 18765 }, { "epoch": 2.387228088029513, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.971989631652832, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8662517070770264, "num_tokens": 716109436.0, "step": 18766 }, { "epoch": 2.3873552983081034, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.887627601623535, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8700577020645142, "num_tokens": 716148157.0, "step": 18767 }, { "epoch": 2.387482508586694, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.963720321655273, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8762396574020386, "num_tokens": 716184662.0, "step": 18768 }, { "epoch": 2.3876097188652845, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.8531494140625, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8667404651641846, "num_tokens": 716227382.0, "step": 18769 }, { "epoch": 2.387736929143875, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.988707542419434, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8777056932449341, "num_tokens": 716262236.0, "step": 18770 }, { "epoch": 2.387864139422465, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.942211151123047, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8824588060379028, "num_tokens": 716295541.0, "step": 18771 }, { "epoch": 2.387991349701056, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.965680122375488, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8764929175376892, "num_tokens": 716331796.0, "step": 18772 }, { "epoch": 2.388118559979646, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.952473640441895, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.864657998085022, "num_tokens": 716369539.0, "step": 18773 }, { "epoch": 2.3882457702582367, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.923842430114746, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.890471339225769, "num_tokens": 716409806.0, "step": 18774 }, { "epoch": 2.3883729805368272, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.913224220275879, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8669562339782715, "num_tokens": 716448191.0, "step": 18775 }, { "epoch": 2.3885001908154178, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.914006233215332, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8754380941390991, "num_tokens": 716486522.0, "step": 18776 }, { "epoch": 2.3886274010940083, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.911890029907227, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8746546506881714, "num_tokens": 716519383.0, "step": 18777 }, { "epoch": 2.388754611372599, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.887406349182129, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8758812546730042, "num_tokens": 716556136.0, "step": 18778 }, { "epoch": 2.3888818216511893, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.91215991973877, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8672739267349243, "num_tokens": 716592346.0, "step": 18779 }, { "epoch": 2.38900903192978, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.951704025268555, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8671237230300903, "num_tokens": 716633996.0, "step": 18780 }, { "epoch": 2.3891362422083704, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.930462837219238, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8698434829711914, "num_tokens": 716669638.0, "step": 18781 }, { "epoch": 2.389263452486961, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.922938346862793, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8650580048561096, "num_tokens": 716711194.0, "step": 18782 }, { "epoch": 2.3893906627655515, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.934701919555664, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8813297748565674, "num_tokens": 716753402.0, "step": 18783 }, { "epoch": 2.389517873044142, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.90468978881836, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8760599493980408, "num_tokens": 716794638.0, "step": 18784 }, { "epoch": 2.3896450833227325, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.968643188476562, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8460586071014404, "num_tokens": 716834003.0, "step": 18785 }, { "epoch": 2.389772293601323, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.931654930114746, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8794912099838257, "num_tokens": 716867349.0, "step": 18786 }, { "epoch": 2.3898995038799136, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.006803512573242, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8724467754364014, "num_tokens": 716904271.0, "step": 18787 }, { "epoch": 2.390026714158504, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.922656059265137, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8763382434844971, "num_tokens": 716936165.0, "step": 18788 }, { "epoch": 2.3901539244370946, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.963184356689453, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.872962236404419, "num_tokens": 716974011.0, "step": 18789 }, { "epoch": 2.390281134715685, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.997590065002441, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8737615346908569, "num_tokens": 717012438.0, "step": 18790 }, { "epoch": 2.3904083449942757, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.941412925720215, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8641810417175293, "num_tokens": 717051755.0, "step": 18791 }, { "epoch": 2.390535555272866, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.920809745788574, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8734159469604492, "num_tokens": 717090740.0, "step": 18792 }, { "epoch": 2.3906627655514567, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.924941062927246, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.848817765712738, "num_tokens": 717132033.0, "step": 18793 }, { "epoch": 2.390789975830047, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.925856590270996, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8809114694595337, "num_tokens": 717168255.0, "step": 18794 }, { "epoch": 2.390917186108638, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.022655487060547, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8721972107887268, "num_tokens": 717206863.0, "step": 18795 }, { "epoch": 2.391044396387228, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.919903755187988, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8531936407089233, "num_tokens": 717245374.0, "step": 18796 }, { "epoch": 2.3911716066658184, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.917200088500977, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8913891315460205, "num_tokens": 717279821.0, "step": 18797 }, { "epoch": 2.391298816944409, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.951112747192383, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8659920692443848, "num_tokens": 717318896.0, "step": 18798 }, { "epoch": 2.3914260272229995, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.926196098327637, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.864302396774292, "num_tokens": 717361493.0, "step": 18799 }, { "epoch": 2.39155323750159, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.96831226348877, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8652403354644775, "num_tokens": 717402801.0, "step": 18800 }, { "epoch": 2.3916804477801805, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.89020824432373, "learning_rate": 1e-06, "loss": 0.3474, "mean_token_accuracy": 0.8909401893615723, "num_tokens": 717441275.0, "step": 18801 }, { "epoch": 2.391807658058771, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.924521446228027, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8611078262329102, "num_tokens": 717477423.0, "step": 18802 }, { "epoch": 2.3919348683373616, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.991105079650879, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8667827844619751, "num_tokens": 717512978.0, "step": 18803 }, { "epoch": 2.392062078615952, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.883888244628906, "learning_rate": 1e-06, "loss": 0.3409, "mean_token_accuracy": 0.8925228714942932, "num_tokens": 717549710.0, "step": 18804 }, { "epoch": 2.3921892888945426, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.969093322753906, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8694281578063965, "num_tokens": 717586588.0, "step": 18805 }, { "epoch": 2.392316499173133, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.913326263427734, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8860101699829102, "num_tokens": 717621507.0, "step": 18806 }, { "epoch": 2.3924437094517237, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.906566619873047, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8753042221069336, "num_tokens": 717655915.0, "step": 18807 }, { "epoch": 2.392570919730314, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.876233100891113, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8706835508346558, "num_tokens": 717696357.0, "step": 18808 }, { "epoch": 2.3926981300089047, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.930490493774414, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8865548372268677, "num_tokens": 717735397.0, "step": 18809 }, { "epoch": 2.3928253402874953, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.987244606018066, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8778601884841919, "num_tokens": 717770081.0, "step": 18810 }, { "epoch": 2.392952550566086, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.96574878692627, "learning_rate": 1e-06, "loss": 0.3435, "mean_token_accuracy": 0.8899235129356384, "num_tokens": 717805262.0, "step": 18811 }, { "epoch": 2.3930797608446763, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.986892700195312, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8738106489181519, "num_tokens": 717841877.0, "step": 18812 }, { "epoch": 2.393206971123267, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.926684379577637, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8581785559654236, "num_tokens": 717881435.0, "step": 18813 }, { "epoch": 2.3933341814018574, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.97360897064209, "learning_rate": 1e-06, "loss": 0.3437, "mean_token_accuracy": 0.8929188847541809, "num_tokens": 717922240.0, "step": 18814 }, { "epoch": 2.393461391680448, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.979799270629883, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8704225420951843, "num_tokens": 717958438.0, "step": 18815 }, { "epoch": 2.3935886019590384, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.918245315551758, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8651671409606934, "num_tokens": 717994451.0, "step": 18816 }, { "epoch": 2.393715812237629, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.918115615844727, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8759279251098633, "num_tokens": 718034318.0, "step": 18817 }, { "epoch": 2.3938430225162195, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.937414169311523, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8727296590805054, "num_tokens": 718074994.0, "step": 18818 }, { "epoch": 2.3939702327948096, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.907404899597168, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8708914518356323, "num_tokens": 718118849.0, "step": 18819 }, { "epoch": 2.3940974430734006, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.92581844329834, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8802158236503601, "num_tokens": 718152874.0, "step": 18820 }, { "epoch": 2.3942246533519906, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.923257827758789, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.866844117641449, "num_tokens": 718194206.0, "step": 18821 }, { "epoch": 2.394351863630581, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.967917442321777, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8764141798019409, "num_tokens": 718227408.0, "step": 18822 }, { "epoch": 2.3944790739091717, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.897588729858398, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8675155639648438, "num_tokens": 718266962.0, "step": 18823 }, { "epoch": 2.3946062841877622, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.937028884887695, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8785650730133057, "num_tokens": 718307873.0, "step": 18824 }, { "epoch": 2.3947334944663528, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.865039825439453, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8793216943740845, "num_tokens": 718350296.0, "step": 18825 }, { "epoch": 2.3948607047449433, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.035776138305664, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.876939058303833, "num_tokens": 718390734.0, "step": 18826 }, { "epoch": 2.394987915023534, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.846190452575684, "learning_rate": 1e-06, "loss": 0.35, "mean_token_accuracy": 0.8893788456916809, "num_tokens": 718435630.0, "step": 18827 }, { "epoch": 2.3951151253021243, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.956721305847168, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8810222148895264, "num_tokens": 718474791.0, "step": 18828 }, { "epoch": 2.395242335580715, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.892707824707031, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8571285009384155, "num_tokens": 718514957.0, "step": 18829 }, { "epoch": 2.3953695458593054, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.989590644836426, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8849917650222778, "num_tokens": 718551761.0, "step": 18830 }, { "epoch": 2.395496756137896, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.928627967834473, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8725259304046631, "num_tokens": 718586445.0, "step": 18831 }, { "epoch": 2.3956239664164865, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.8626708984375, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8689240217208862, "num_tokens": 718631105.0, "step": 18832 }, { "epoch": 2.395751176695077, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.964069366455078, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8699225187301636, "num_tokens": 718672278.0, "step": 18833 }, { "epoch": 2.3958783869736675, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.947625160217285, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8794423937797546, "num_tokens": 718713191.0, "step": 18834 }, { "epoch": 2.396005597252258, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.957785606384277, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8675810098648071, "num_tokens": 718745012.0, "step": 18835 }, { "epoch": 2.3961328075308486, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.907254219055176, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8739470839500427, "num_tokens": 718780149.0, "step": 18836 }, { "epoch": 2.396260017809439, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.941627502441406, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8693497776985168, "num_tokens": 718812831.0, "step": 18837 }, { "epoch": 2.3963872280880296, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.023874282836914, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8839938640594482, "num_tokens": 718847650.0, "step": 18838 }, { "epoch": 2.39651443836662, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.868297576904297, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8749442100524902, "num_tokens": 718883417.0, "step": 18839 }, { "epoch": 2.3966416486452107, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.894769668579102, "learning_rate": 1e-06, "loss": 0.3175, "mean_token_accuracy": 0.897484540939331, "num_tokens": 718915501.0, "step": 18840 }, { "epoch": 2.396768858923801, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.955362319946289, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8883048892021179, "num_tokens": 718954485.0, "step": 18841 }, { "epoch": 2.3968960692023917, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.92788314819336, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8590630292892456, "num_tokens": 718995815.0, "step": 18842 }, { "epoch": 2.3970232794809823, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.887237548828125, "learning_rate": 1e-06, "loss": 0.336, "mean_token_accuracy": 0.8877564072608948, "num_tokens": 719032820.0, "step": 18843 }, { "epoch": 2.3971504897595723, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.974343299865723, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8711502552032471, "num_tokens": 719066966.0, "step": 18844 }, { "epoch": 2.3972777000381633, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.93784236907959, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8734651803970337, "num_tokens": 719108768.0, "step": 18845 }, { "epoch": 2.3974049103167534, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.943365097045898, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8709560036659241, "num_tokens": 719151451.0, "step": 18846 }, { "epoch": 2.397532120595344, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.945745468139648, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.866157054901123, "num_tokens": 719192103.0, "step": 18847 }, { "epoch": 2.3976593308739345, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.992721557617188, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8708683848381042, "num_tokens": 719226514.0, "step": 18848 }, { "epoch": 2.397786541152525, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.98819637298584, "learning_rate": 1e-06, "loss": 0.3242, "mean_token_accuracy": 0.8970522284507751, "num_tokens": 719262944.0, "step": 18849 }, { "epoch": 2.3979137514311155, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.939677238464355, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8796278238296509, "num_tokens": 719306707.0, "step": 18850 }, { "epoch": 2.398040961709706, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.959507942199707, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8789805173873901, "num_tokens": 719347523.0, "step": 18851 }, { "epoch": 2.3981681719882966, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.005901336669922, "learning_rate": 1e-06, "loss": 0.3291, "mean_token_accuracy": 0.8931399583816528, "num_tokens": 719379474.0, "step": 18852 }, { "epoch": 2.398295382266887, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.922794342041016, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8588638305664062, "num_tokens": 719421624.0, "step": 18853 }, { "epoch": 2.3984225925454776, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.049243927001953, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8692846894264221, "num_tokens": 719455570.0, "step": 18854 }, { "epoch": 2.398549802824068, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.90710163116455, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8907418251037598, "num_tokens": 719500300.0, "step": 18855 }, { "epoch": 2.3986770131026587, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.962884902954102, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8588404059410095, "num_tokens": 719538119.0, "step": 18856 }, { "epoch": 2.398804223381249, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.035908699035645, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8651655912399292, "num_tokens": 719574012.0, "step": 18857 }, { "epoch": 2.3989314336598397, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.935127258300781, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8854414224624634, "num_tokens": 719609548.0, "step": 18858 }, { "epoch": 2.3990586439384303, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.959593772888184, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8677590489387512, "num_tokens": 719647280.0, "step": 18859 }, { "epoch": 2.399185854217021, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.942924499511719, "learning_rate": 1e-06, "loss": 0.3482, "mean_token_accuracy": 0.8890421390533447, "num_tokens": 719686987.0, "step": 18860 }, { "epoch": 2.3993130644956113, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.87863826751709, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8921623229980469, "num_tokens": 719727974.0, "step": 18861 }, { "epoch": 2.399440274774202, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.990533828735352, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8747286796569824, "num_tokens": 719764308.0, "step": 18862 }, { "epoch": 2.3995674850527924, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.979289054870605, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8642356991767883, "num_tokens": 719801872.0, "step": 18863 }, { "epoch": 2.399694695331383, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.996392250061035, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8768140077590942, "num_tokens": 719842504.0, "step": 18864 }, { "epoch": 2.3998219056099734, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.979714393615723, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8686308860778809, "num_tokens": 719878885.0, "step": 18865 }, { "epoch": 2.399949115888564, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.988018989562988, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8573275804519653, "num_tokens": 719922891.0, "step": 18866 }, { "epoch": 2.4000763261671545, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.020865440368652, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8725464940071106, "num_tokens": 719960113.0, "step": 18867 }, { "epoch": 2.400203536445745, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.929764747619629, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8731417655944824, "num_tokens": 720006079.0, "step": 18868 }, { "epoch": 2.400330746724335, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.925366401672363, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8613208532333374, "num_tokens": 720042219.0, "step": 18869 }, { "epoch": 2.400457957002926, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.932174682617188, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8748151659965515, "num_tokens": 720082574.0, "step": 18870 }, { "epoch": 2.400585167281516, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.990788459777832, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8721064925193787, "num_tokens": 720115583.0, "step": 18871 }, { "epoch": 2.4007123775601067, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.903457641601562, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8654873371124268, "num_tokens": 720151350.0, "step": 18872 }, { "epoch": 2.400839587838697, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.919126510620117, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8660119771957397, "num_tokens": 720187072.0, "step": 18873 }, { "epoch": 2.4009667981172877, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.920402526855469, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8777512311935425, "num_tokens": 720222249.0, "step": 18874 }, { "epoch": 2.4010940083958783, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.952993392944336, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8724378943443298, "num_tokens": 720264682.0, "step": 18875 }, { "epoch": 2.401221218674469, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.898799896240234, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8707593083381653, "num_tokens": 720309935.0, "step": 18876 }, { "epoch": 2.4013484289530593, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.98819351196289, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8690902590751648, "num_tokens": 720343992.0, "step": 18877 }, { "epoch": 2.40147563923165, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.940812110900879, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.873809278011322, "num_tokens": 720379707.0, "step": 18878 }, { "epoch": 2.4016028495102404, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.972593307495117, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8702253103256226, "num_tokens": 720419796.0, "step": 18879 }, { "epoch": 2.401730059788831, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.914040565490723, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8757764101028442, "num_tokens": 720460919.0, "step": 18880 }, { "epoch": 2.4018572700674214, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.952324867248535, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8696746826171875, "num_tokens": 720498403.0, "step": 18881 }, { "epoch": 2.401984480346012, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.971281051635742, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.874764621257782, "num_tokens": 720534931.0, "step": 18882 }, { "epoch": 2.4021116906246025, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.982389450073242, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8766265511512756, "num_tokens": 720574062.0, "step": 18883 }, { "epoch": 2.402238900903193, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.992096900939941, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8660266399383545, "num_tokens": 720611552.0, "step": 18884 }, { "epoch": 2.4023661111817836, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.981471061706543, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8728251457214355, "num_tokens": 720647886.0, "step": 18885 }, { "epoch": 2.402493321460374, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.99160099029541, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8793309926986694, "num_tokens": 720688034.0, "step": 18886 }, { "epoch": 2.4026205317389646, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.966729164123535, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8805615901947021, "num_tokens": 720727805.0, "step": 18887 }, { "epoch": 2.402747742017555, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.06440544128418, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.876772403717041, "num_tokens": 720759717.0, "step": 18888 }, { "epoch": 2.4028749522961457, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.95961856842041, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.867698073387146, "num_tokens": 720790435.0, "step": 18889 }, { "epoch": 2.403002162574736, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.025118827819824, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8692996501922607, "num_tokens": 720831715.0, "step": 18890 }, { "epoch": 2.4031293728533267, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.923768043518066, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8724746704101562, "num_tokens": 720869598.0, "step": 18891 }, { "epoch": 2.403256583131917, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.032902717590332, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.868969738483429, "num_tokens": 720911122.0, "step": 18892 }, { "epoch": 2.403383793410508, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.960152626037598, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8771759867668152, "num_tokens": 720952265.0, "step": 18893 }, { "epoch": 2.403511003689098, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.99052619934082, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8771542310714722, "num_tokens": 720991191.0, "step": 18894 }, { "epoch": 2.4036382139676884, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.944113731384277, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8839129209518433, "num_tokens": 721032524.0, "step": 18895 }, { "epoch": 2.403765424246279, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.949957847595215, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8736225366592407, "num_tokens": 721072699.0, "step": 18896 }, { "epoch": 2.4038926345248695, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.949363708496094, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8856897354125977, "num_tokens": 721109467.0, "step": 18897 }, { "epoch": 2.40401984480346, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.029086112976074, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8505399227142334, "num_tokens": 721142108.0, "step": 18898 }, { "epoch": 2.4041470550820505, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.927583694458008, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8671042919158936, "num_tokens": 721179875.0, "step": 18899 }, { "epoch": 2.404274265360641, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.961114883422852, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8708631992340088, "num_tokens": 721217141.0, "step": 18900 }, { "epoch": 2.4044014756392316, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.92894458770752, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8677834272384644, "num_tokens": 721256357.0, "step": 18901 }, { "epoch": 2.404528685917822, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.9351224899292, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8735391497612, "num_tokens": 721292610.0, "step": 18902 }, { "epoch": 2.4046558961964126, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.968071937561035, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8796051740646362, "num_tokens": 721329048.0, "step": 18903 }, { "epoch": 2.404783106475003, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.969511032104492, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8893477320671082, "num_tokens": 721364249.0, "step": 18904 }, { "epoch": 2.4049103167535937, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.992913246154785, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8803786039352417, "num_tokens": 721391154.0, "step": 18905 }, { "epoch": 2.405037527032184, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.006926536560059, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8840711712837219, "num_tokens": 721428443.0, "step": 18906 }, { "epoch": 2.4051647373107747, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.986702919006348, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8641743063926697, "num_tokens": 721474545.0, "step": 18907 }, { "epoch": 2.4052919475893653, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.943286895751953, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8743548393249512, "num_tokens": 721513153.0, "step": 18908 }, { "epoch": 2.405419157867956, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.973177909851074, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8701512217521667, "num_tokens": 721556287.0, "step": 18909 }, { "epoch": 2.4055463681465463, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.956268310546875, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8800649642944336, "num_tokens": 721590219.0, "step": 18910 }, { "epoch": 2.405673578425137, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.95310115814209, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8730611801147461, "num_tokens": 721627308.0, "step": 18911 }, { "epoch": 2.4058007887037274, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.047345161437988, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8695468902587891, "num_tokens": 721664218.0, "step": 18912 }, { "epoch": 2.405927998982318, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.90206241607666, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8850244283676147, "num_tokens": 721703511.0, "step": 18913 }, { "epoch": 2.4060552092609084, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.985357284545898, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8857058882713318, "num_tokens": 721741512.0, "step": 18914 }, { "epoch": 2.406182419539499, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.990509986877441, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.877217173576355, "num_tokens": 721778443.0, "step": 18915 }, { "epoch": 2.4063096298180895, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.950705528259277, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8702172040939331, "num_tokens": 721817226.0, "step": 18916 }, { "epoch": 2.4064368400966796, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.9710111618042, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8600454330444336, "num_tokens": 721859419.0, "step": 18917 }, { "epoch": 2.4065640503752705, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.931159019470215, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8866580724716187, "num_tokens": 721895183.0, "step": 18918 }, { "epoch": 2.4066912606538606, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.939657211303711, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.859703779220581, "num_tokens": 721933566.0, "step": 18919 }, { "epoch": 2.406818470932451, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.981115341186523, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8779271841049194, "num_tokens": 721968827.0, "step": 18920 }, { "epoch": 2.4069456812110417, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.88863754272461, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8768649101257324, "num_tokens": 722004989.0, "step": 18921 }, { "epoch": 2.407072891489632, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.014352798461914, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8766670823097229, "num_tokens": 722043681.0, "step": 18922 }, { "epoch": 2.4072001017682227, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.878397941589355, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8735334873199463, "num_tokens": 722086541.0, "step": 18923 }, { "epoch": 2.4073273120468133, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.946287155151367, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8743208646774292, "num_tokens": 722125519.0, "step": 18924 }, { "epoch": 2.407454522325404, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.976710319519043, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8699523210525513, "num_tokens": 722157665.0, "step": 18925 }, { "epoch": 2.4075817326039943, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.000724792480469, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8862184286117554, "num_tokens": 722191819.0, "step": 18926 }, { "epoch": 2.407708942882585, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.046887397766113, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8633953332901001, "num_tokens": 722229755.0, "step": 18927 }, { "epoch": 2.4078361531611754, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.948931694030762, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8627111911773682, "num_tokens": 722267440.0, "step": 18928 }, { "epoch": 2.407963363439766, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.001799583435059, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8694985508918762, "num_tokens": 722310102.0, "step": 18929 }, { "epoch": 2.4080905737183564, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.999890327453613, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8672798275947571, "num_tokens": 722346713.0, "step": 18930 }, { "epoch": 2.408217783996947, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.980480194091797, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8685509562492371, "num_tokens": 722387803.0, "step": 18931 }, { "epoch": 2.4083449942755375, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.983874320983887, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.865709662437439, "num_tokens": 722428943.0, "step": 18932 }, { "epoch": 2.408472204554128, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.015817642211914, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.867800235748291, "num_tokens": 722467860.0, "step": 18933 }, { "epoch": 2.4085994148327186, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.914217948913574, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.868931233882904, "num_tokens": 722503067.0, "step": 18934 }, { "epoch": 2.408726625111309, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.969061851501465, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8763487935066223, "num_tokens": 722537842.0, "step": 18935 }, { "epoch": 2.4088538353898996, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.949063301086426, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8798198103904724, "num_tokens": 722574036.0, "step": 18936 }, { "epoch": 2.40898104566849, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.9324369430542, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8743232488632202, "num_tokens": 722606087.0, "step": 18937 }, { "epoch": 2.4091082559470807, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.995627403259277, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8723131418228149, "num_tokens": 722648745.0, "step": 18938 }, { "epoch": 2.409235466225671, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.923698425292969, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8656744956970215, "num_tokens": 722688589.0, "step": 18939 }, { "epoch": 2.4093626765042617, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.975903511047363, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8905283212661743, "num_tokens": 722723941.0, "step": 18940 }, { "epoch": 2.4094898867828523, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.97309684753418, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8568243980407715, "num_tokens": 722768242.0, "step": 18941 }, { "epoch": 2.4096170970614423, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.004800796508789, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8738878965377808, "num_tokens": 722800970.0, "step": 18942 }, { "epoch": 2.4097443073400333, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.972081184387207, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8643385171890259, "num_tokens": 722839168.0, "step": 18943 }, { "epoch": 2.4098715176186234, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.991767883300781, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8849617838859558, "num_tokens": 722883451.0, "step": 18944 }, { "epoch": 2.409998727897214, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.00687026977539, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8747819662094116, "num_tokens": 722928581.0, "step": 18945 }, { "epoch": 2.4101259381758044, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.948107719421387, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8847073912620544, "num_tokens": 722967141.0, "step": 18946 }, { "epoch": 2.410253148454395, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.979235649108887, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8697319030761719, "num_tokens": 723006263.0, "step": 18947 }, { "epoch": 2.4103803587329855, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.947930335998535, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8768094778060913, "num_tokens": 723050843.0, "step": 18948 }, { "epoch": 2.410507569011576, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.988547325134277, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8739764094352722, "num_tokens": 723092124.0, "step": 18949 }, { "epoch": 2.4106347792901666, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.969653129577637, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8850187063217163, "num_tokens": 723122478.0, "step": 18950 }, { "epoch": 2.410761989568757, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.974296569824219, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8680934309959412, "num_tokens": 723163054.0, "step": 18951 }, { "epoch": 2.4108891998473476, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.995316505432129, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8731083869934082, "num_tokens": 723201518.0, "step": 18952 }, { "epoch": 2.411016410125938, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.960288047790527, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8740712404251099, "num_tokens": 723235685.0, "step": 18953 }, { "epoch": 2.4111436204045287, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.048186302185059, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8737442493438721, "num_tokens": 723266510.0, "step": 18954 }, { "epoch": 2.411270830683119, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.001738548278809, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8771542310714722, "num_tokens": 723302800.0, "step": 18955 }, { "epoch": 2.4113980409617097, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.976217269897461, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8763138651847839, "num_tokens": 723340669.0, "step": 18956 }, { "epoch": 2.4115252512403003, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.052891731262207, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8768382668495178, "num_tokens": 723377556.0, "step": 18957 }, { "epoch": 2.411652461518891, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.927552223205566, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8791091442108154, "num_tokens": 723410942.0, "step": 18958 }, { "epoch": 2.4117796717974813, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.970802307128906, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8868794441223145, "num_tokens": 723451131.0, "step": 18959 }, { "epoch": 2.411906882076072, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.983651161193848, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.877430260181427, "num_tokens": 723485876.0, "step": 18960 }, { "epoch": 2.4120340923546624, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.997335433959961, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8819725513458252, "num_tokens": 723521252.0, "step": 18961 }, { "epoch": 2.412161302633253, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.93864631652832, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8586862683296204, "num_tokens": 723561094.0, "step": 18962 }, { "epoch": 2.4122885129118434, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.022832870483398, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8726403117179871, "num_tokens": 723597281.0, "step": 18963 }, { "epoch": 2.412415723190434, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.947402000427246, "learning_rate": 1e-06, "loss": 0.3251, "mean_token_accuracy": 0.8967593908309937, "num_tokens": 723637690.0, "step": 18964 }, { "epoch": 2.4125429334690245, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.990019798278809, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8746268153190613, "num_tokens": 723676250.0, "step": 18965 }, { "epoch": 2.412670143747615, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.943168640136719, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8829613924026489, "num_tokens": 723713157.0, "step": 18966 }, { "epoch": 2.412797354026205, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.961153030395508, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8793984651565552, "num_tokens": 723745877.0, "step": 18967 }, { "epoch": 2.412924564304796, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.936392784118652, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8752771615982056, "num_tokens": 723784285.0, "step": 18968 }, { "epoch": 2.413051774583386, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.072718620300293, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8625695705413818, "num_tokens": 723817085.0, "step": 18969 }, { "epoch": 2.4131789848619767, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.933679580688477, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8695787191390991, "num_tokens": 723858466.0, "step": 18970 }, { "epoch": 2.413306195140567, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.915202140808105, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8733729124069214, "num_tokens": 723896627.0, "step": 18971 }, { "epoch": 2.4134334054191577, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.052520751953125, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8693201541900635, "num_tokens": 723930727.0, "step": 18972 }, { "epoch": 2.4135606156977483, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.918488502502441, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8641211986541748, "num_tokens": 723965996.0, "step": 18973 }, { "epoch": 2.413687825976339, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.043516159057617, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.869049072265625, "num_tokens": 724002441.0, "step": 18974 }, { "epoch": 2.4138150362549293, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.961440086364746, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8782256245613098, "num_tokens": 724039238.0, "step": 18975 }, { "epoch": 2.41394224653352, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.911067008972168, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8790427446365356, "num_tokens": 724077430.0, "step": 18976 }, { "epoch": 2.4140694568121104, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.92618465423584, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8706338405609131, "num_tokens": 724113175.0, "step": 18977 }, { "epoch": 2.414196667090701, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.925801277160645, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8692416548728943, "num_tokens": 724156409.0, "step": 18978 }, { "epoch": 2.4143238773692914, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.935290336608887, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8743677139282227, "num_tokens": 724199879.0, "step": 18979 }, { "epoch": 2.414451087647882, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.992228507995605, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8865991234779358, "num_tokens": 724238065.0, "step": 18980 }, { "epoch": 2.4145782979264725, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.998245239257812, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8673251271247864, "num_tokens": 724272336.0, "step": 18981 }, { "epoch": 2.414705508205063, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.943493843078613, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8657112121582031, "num_tokens": 724305200.0, "step": 18982 }, { "epoch": 2.4148327184836536, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.987634658813477, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8787894248962402, "num_tokens": 724341295.0, "step": 18983 }, { "epoch": 2.414959928762244, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.970212936401367, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8763300180435181, "num_tokens": 724384336.0, "step": 18984 }, { "epoch": 2.4150871390408346, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.985640525817871, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8709304928779602, "num_tokens": 724417279.0, "step": 18985 }, { "epoch": 2.415214349319425, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.912457466125488, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8732787370681763, "num_tokens": 724456509.0, "step": 18986 }, { "epoch": 2.4153415595980157, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.96418285369873, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8841234445571899, "num_tokens": 724498812.0, "step": 18987 }, { "epoch": 2.415468769876606, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.950142860412598, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8621578812599182, "num_tokens": 724538174.0, "step": 18988 }, { "epoch": 2.4155959801551967, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.955968856811523, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8800018429756165, "num_tokens": 724577171.0, "step": 18989 }, { "epoch": 2.415723190433787, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.010472297668457, "learning_rate": 1e-06, "loss": 0.3324, "mean_token_accuracy": 0.8952834010124207, "num_tokens": 724610626.0, "step": 18990 }, { "epoch": 2.4158504007123778, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.927434921264648, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8856918811798096, "num_tokens": 724647613.0, "step": 18991 }, { "epoch": 2.415977610990968, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.97680950164795, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8759839534759521, "num_tokens": 724684898.0, "step": 18992 }, { "epoch": 2.4161048212695584, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.027937889099121, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8703730702400208, "num_tokens": 724725566.0, "step": 18993 }, { "epoch": 2.416232031548149, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.941630363464355, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8651129007339478, "num_tokens": 724763701.0, "step": 18994 }, { "epoch": 2.4163592418267394, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.942224502563477, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.873816192150116, "num_tokens": 724798108.0, "step": 18995 }, { "epoch": 2.41648645210533, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.960206985473633, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8541965484619141, "num_tokens": 724842430.0, "step": 18996 }, { "epoch": 2.4166136623839205, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.981138229370117, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8702786564826965, "num_tokens": 724884585.0, "step": 18997 }, { "epoch": 2.416740872662511, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.97256088256836, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8853211402893066, "num_tokens": 724919338.0, "step": 18998 }, { "epoch": 2.4168680829411016, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.897907257080078, "learning_rate": 1e-06, "loss": 0.3353, "mean_token_accuracy": 0.8952909708023071, "num_tokens": 724956007.0, "step": 18999 }, { "epoch": 2.416995293219692, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.996048927307129, "learning_rate": 1e-06, "loss": 0.3486, "mean_token_accuracy": 0.8903120756149292, "num_tokens": 724996994.0, "step": 19000 }, { "epoch": 2.4171225034982826, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.963665962219238, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8679718971252441, "num_tokens": 725035607.0, "step": 19001 }, { "epoch": 2.417249713776873, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.910760879516602, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8770447373390198, "num_tokens": 725075370.0, "step": 19002 }, { "epoch": 2.4173769240554637, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.04306697845459, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8717600703239441, "num_tokens": 725119536.0, "step": 19003 }, { "epoch": 2.417504134334054, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.926795959472656, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8736774921417236, "num_tokens": 725158048.0, "step": 19004 }, { "epoch": 2.4176313446126447, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.966585159301758, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8829212784767151, "num_tokens": 725196343.0, "step": 19005 }, { "epoch": 2.4177585548912353, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.091796875, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8743848204612732, "num_tokens": 725234613.0, "step": 19006 }, { "epoch": 2.417885765169826, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.997426986694336, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8727449178695679, "num_tokens": 725272129.0, "step": 19007 }, { "epoch": 2.4180129754484163, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.07760238647461, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8647996783256531, "num_tokens": 725305740.0, "step": 19008 }, { "epoch": 2.418140185727007, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.013521194458008, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8710045218467712, "num_tokens": 725340579.0, "step": 19009 }, { "epoch": 2.4182673960055974, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.970979690551758, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8768166303634644, "num_tokens": 725380492.0, "step": 19010 }, { "epoch": 2.418394606284188, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.952157020568848, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8767188787460327, "num_tokens": 725418043.0, "step": 19011 }, { "epoch": 2.4185218165627784, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.052251815795898, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8655409216880798, "num_tokens": 725456260.0, "step": 19012 }, { "epoch": 2.418649026841369, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.951608657836914, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8708703517913818, "num_tokens": 725496605.0, "step": 19013 }, { "epoch": 2.4187762371199595, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.027878761291504, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8760504722595215, "num_tokens": 725532395.0, "step": 19014 }, { "epoch": 2.4189034473985496, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.947159767150879, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8688618540763855, "num_tokens": 725575335.0, "step": 19015 }, { "epoch": 2.4190306576771405, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.037212371826172, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8806458711624146, "num_tokens": 725612345.0, "step": 19016 }, { "epoch": 2.4191578679557306, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.02875804901123, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8774564266204834, "num_tokens": 725645566.0, "step": 19017 }, { "epoch": 2.419285078234321, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.956998825073242, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8684322834014893, "num_tokens": 725684218.0, "step": 19018 }, { "epoch": 2.4194122885129117, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.01579761505127, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8779894709587097, "num_tokens": 725718478.0, "step": 19019 }, { "epoch": 2.419539498791502, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.041072845458984, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8676888942718506, "num_tokens": 725756208.0, "step": 19020 }, { "epoch": 2.4196667090700927, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.008492469787598, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8681684732437134, "num_tokens": 725791884.0, "step": 19021 }, { "epoch": 2.4197939193486833, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.021405220031738, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8777642250061035, "num_tokens": 725829138.0, "step": 19022 }, { "epoch": 2.419921129627274, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.979491233825684, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8745085597038269, "num_tokens": 725868301.0, "step": 19023 }, { "epoch": 2.4200483399058643, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.017843246459961, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8635079264640808, "num_tokens": 725906186.0, "step": 19024 }, { "epoch": 2.420175550184455, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.087674140930176, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8621951341629028, "num_tokens": 725947006.0, "step": 19025 }, { "epoch": 2.4203027604630454, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.041439056396484, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8777439594268799, "num_tokens": 725987363.0, "step": 19026 }, { "epoch": 2.420429970741636, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.039133071899414, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8646301627159119, "num_tokens": 726026176.0, "step": 19027 }, { "epoch": 2.4205571810202264, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.994735717773438, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8724863529205322, "num_tokens": 726068124.0, "step": 19028 }, { "epoch": 2.420684391298817, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.990335464477539, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8656604290008545, "num_tokens": 726110270.0, "step": 19029 }, { "epoch": 2.4208116015774075, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.045073509216309, "learning_rate": 1e-06, "loss": 0.346, "mean_token_accuracy": 0.889254093170166, "num_tokens": 726143908.0, "step": 19030 }, { "epoch": 2.420938811855998, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.093168258666992, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8718751668930054, "num_tokens": 726179082.0, "step": 19031 }, { "epoch": 2.4210660221345885, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.017890930175781, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8673197031021118, "num_tokens": 726216387.0, "step": 19032 }, { "epoch": 2.421193232413179, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.045341491699219, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8831114172935486, "num_tokens": 726258831.0, "step": 19033 }, { "epoch": 2.4213204426917696, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.074317932128906, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8754637241363525, "num_tokens": 726301929.0, "step": 19034 }, { "epoch": 2.42144765297036, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.025049209594727, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8768514394760132, "num_tokens": 726339938.0, "step": 19035 }, { "epoch": 2.4215748632489507, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.042868614196777, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8748962879180908, "num_tokens": 726377197.0, "step": 19036 }, { "epoch": 2.421702073527541, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.079672813415527, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8779628276824951, "num_tokens": 726417591.0, "step": 19037 }, { "epoch": 2.4218292838061317, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.013331413269043, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8642621040344238, "num_tokens": 726456094.0, "step": 19038 }, { "epoch": 2.4219564940847222, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.085699081420898, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8833906650543213, "num_tokens": 726491158.0, "step": 19039 }, { "epoch": 2.4220837043633123, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.038678169250488, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8601399660110474, "num_tokens": 726531334.0, "step": 19040 }, { "epoch": 2.4222109146419033, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.969402313232422, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8647469282150269, "num_tokens": 726571999.0, "step": 19041 }, { "epoch": 2.4223381249204934, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.135933876037598, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8644957542419434, "num_tokens": 726604945.0, "step": 19042 }, { "epoch": 2.422465335199084, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.073299407958984, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8654568195343018, "num_tokens": 726645023.0, "step": 19043 }, { "epoch": 2.4225925454776744, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.016990661621094, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8724944591522217, "num_tokens": 726684297.0, "step": 19044 }, { "epoch": 2.422719755756265, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.0125732421875, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.86350417137146, "num_tokens": 726726950.0, "step": 19045 }, { "epoch": 2.4228469660348555, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.059542655944824, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8549596071243286, "num_tokens": 726761944.0, "step": 19046 }, { "epoch": 2.422974176313446, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.062891960144043, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8835614323616028, "num_tokens": 726799041.0, "step": 19047 }, { "epoch": 2.4231013865920366, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.018254280090332, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8682090640068054, "num_tokens": 726838160.0, "step": 19048 }, { "epoch": 2.423228596870627, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.978534698486328, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.883699893951416, "num_tokens": 726873150.0, "step": 19049 }, { "epoch": 2.4233558071492176, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.987316131591797, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8772227764129639, "num_tokens": 726905253.0, "step": 19050 }, { "epoch": 2.423483017427808, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.119657516479492, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8835009336471558, "num_tokens": 726942842.0, "step": 19051 }, { "epoch": 2.4236102277063987, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.948065757751465, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8775262832641602, "num_tokens": 726987974.0, "step": 19052 }, { "epoch": 2.423737437984989, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.021129608154297, "learning_rate": 1e-06, "loss": 0.3435, "mean_token_accuracy": 0.8922007083892822, "num_tokens": 727021614.0, "step": 19053 }, { "epoch": 2.4238646482635797, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.069157600402832, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8720742464065552, "num_tokens": 727061773.0, "step": 19054 }, { "epoch": 2.4239918585421703, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.021367073059082, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8687057495117188, "num_tokens": 727095877.0, "step": 19055 }, { "epoch": 2.424119068820761, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.049213409423828, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8767604231834412, "num_tokens": 727128676.0, "step": 19056 }, { "epoch": 2.4242462790993513, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.03409481048584, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8827469348907471, "num_tokens": 727164528.0, "step": 19057 }, { "epoch": 2.424373489377942, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.997798919677734, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8805283308029175, "num_tokens": 727202403.0, "step": 19058 }, { "epoch": 2.4245006996565324, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.080233573913574, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8768103122711182, "num_tokens": 727242231.0, "step": 19059 }, { "epoch": 2.424627909935123, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.06519889831543, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8613367676734924, "num_tokens": 727276371.0, "step": 19060 }, { "epoch": 2.4247551202137134, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.955971717834473, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8776459693908691, "num_tokens": 727320101.0, "step": 19061 }, { "epoch": 2.424882330492304, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.99696159362793, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8651728630065918, "num_tokens": 727363687.0, "step": 19062 }, { "epoch": 2.4250095407708945, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.977372169494629, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8794628381729126, "num_tokens": 727399772.0, "step": 19063 }, { "epoch": 2.425136751049485, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.104899406433105, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8818234205245972, "num_tokens": 727442944.0, "step": 19064 }, { "epoch": 2.425263961328075, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.94857120513916, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8823380470275879, "num_tokens": 727483283.0, "step": 19065 }, { "epoch": 2.425391171606666, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.093490600585938, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8725936412811279, "num_tokens": 727526316.0, "step": 19066 }, { "epoch": 2.425518381885256, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.016751289367676, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8805391788482666, "num_tokens": 727563218.0, "step": 19067 }, { "epoch": 2.4256455921638467, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.05638599395752, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8867173194885254, "num_tokens": 727594877.0, "step": 19068 }, { "epoch": 2.425772802442437, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.127158164978027, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8870693445205688, "num_tokens": 727627561.0, "step": 19069 }, { "epoch": 2.4259000127210277, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.050113677978516, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.865415096282959, "num_tokens": 727664384.0, "step": 19070 }, { "epoch": 2.4260272229996183, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.99889850616455, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8712915778160095, "num_tokens": 727700866.0, "step": 19071 }, { "epoch": 2.426154433278209, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.09228515625, "learning_rate": 1e-06, "loss": 0.3502, "mean_token_accuracy": 0.8912977576255798, "num_tokens": 727741539.0, "step": 19072 }, { "epoch": 2.4262816435567993, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.9889497756958, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8750706315040588, "num_tokens": 727777828.0, "step": 19073 }, { "epoch": 2.42640885383539, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.05960464477539, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8816473484039307, "num_tokens": 727818979.0, "step": 19074 }, { "epoch": 2.4265360641139804, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.98344898223877, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8756418824195862, "num_tokens": 727856440.0, "step": 19075 }, { "epoch": 2.426663274392571, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.037216186523438, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8822355270385742, "num_tokens": 727895036.0, "step": 19076 }, { "epoch": 2.4267904846711614, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.091373443603516, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.863914430141449, "num_tokens": 727932292.0, "step": 19077 }, { "epoch": 2.426917694949752, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.051776885986328, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.874726414680481, "num_tokens": 727967005.0, "step": 19078 }, { "epoch": 2.4270449052283425, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.03017807006836, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8650654554367065, "num_tokens": 728004262.0, "step": 19079 }, { "epoch": 2.427172115506933, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.9978666305542, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8838508725166321, "num_tokens": 728041451.0, "step": 19080 }, { "epoch": 2.4272993257855235, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.0630521774292, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8723176717758179, "num_tokens": 728079217.0, "step": 19081 }, { "epoch": 2.427426536064114, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.985662460327148, "learning_rate": 1e-06, "loss": 0.336, "mean_token_accuracy": 0.8904333114624023, "num_tokens": 728116221.0, "step": 19082 }, { "epoch": 2.4275537463427046, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.048648834228516, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8666343092918396, "num_tokens": 728154861.0, "step": 19083 }, { "epoch": 2.427680956621295, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.014415740966797, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8639906048774719, "num_tokens": 728188276.0, "step": 19084 }, { "epoch": 2.4278081668998857, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.014167785644531, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8853396773338318, "num_tokens": 728229313.0, "step": 19085 }, { "epoch": 2.427935377178476, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.03690242767334, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8632665276527405, "num_tokens": 728268706.0, "step": 19086 }, { "epoch": 2.4280625874570667, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.074323654174805, "learning_rate": 1e-06, "loss": 0.3502, "mean_token_accuracy": 0.886383056640625, "num_tokens": 728297490.0, "step": 19087 }, { "epoch": 2.428189797735657, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.024322509765625, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8669807314872742, "num_tokens": 728336041.0, "step": 19088 }, { "epoch": 2.4283170080142478, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.036118507385254, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8811664581298828, "num_tokens": 728372701.0, "step": 19089 }, { "epoch": 2.428444218292838, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.08787727355957, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8764657974243164, "num_tokens": 728408138.0, "step": 19090 }, { "epoch": 2.4285714285714284, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.01246166229248, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8808237314224243, "num_tokens": 728452649.0, "step": 19091 }, { "epoch": 2.428698638850019, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.070534706115723, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8649551868438721, "num_tokens": 728490551.0, "step": 19092 }, { "epoch": 2.4288258491286094, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.03310775756836, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8804799318313599, "num_tokens": 728529558.0, "step": 19093 }, { "epoch": 2.4289530594072, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.94748592376709, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8518691658973694, "num_tokens": 728573495.0, "step": 19094 }, { "epoch": 2.4290802696857905, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.083457946777344, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8757282495498657, "num_tokens": 728609166.0, "step": 19095 }, { "epoch": 2.429207479964381, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.022406578063965, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8808982372283936, "num_tokens": 728653070.0, "step": 19096 }, { "epoch": 2.4293346902429716, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.977330207824707, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8773093819618225, "num_tokens": 728689142.0, "step": 19097 }, { "epoch": 2.429461900521562, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.091791152954102, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8792980909347534, "num_tokens": 728725615.0, "step": 19098 }, { "epoch": 2.4295891108001526, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.017423629760742, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8651280999183655, "num_tokens": 728762366.0, "step": 19099 }, { "epoch": 2.429716321078743, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.97503662109375, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8622936010360718, "num_tokens": 728801951.0, "step": 19100 }, { "epoch": 2.4298435313573337, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.070799827575684, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8826486468315125, "num_tokens": 728836403.0, "step": 19101 }, { "epoch": 2.429970741635924, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.014666557312012, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8813716173171997, "num_tokens": 728878485.0, "step": 19102 }, { "epoch": 2.4300979519145147, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.96855640411377, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8835997581481934, "num_tokens": 728915935.0, "step": 19103 }, { "epoch": 2.4302251621931052, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.003488540649414, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8578728437423706, "num_tokens": 728956847.0, "step": 19104 }, { "epoch": 2.4303523724716958, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.999762535095215, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8801275491714478, "num_tokens": 728994865.0, "step": 19105 }, { "epoch": 2.4304795827502863, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.016688346862793, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8677773475646973, "num_tokens": 729033030.0, "step": 19106 }, { "epoch": 2.430606793028877, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.981063842773438, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8851220607757568, "num_tokens": 729070430.0, "step": 19107 }, { "epoch": 2.4307340033074674, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 14.046865463256836, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8851836323738098, "num_tokens": 729102127.0, "step": 19108 }, { "epoch": 2.430861213586058, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.993191719055176, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8775056004524231, "num_tokens": 729138128.0, "step": 19109 }, { "epoch": 2.4309884238646484, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.98740005493164, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8718280792236328, "num_tokens": 729179756.0, "step": 19110 }, { "epoch": 2.431115634143239, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.94494915008545, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8633294105529785, "num_tokens": 729221356.0, "step": 19111 }, { "epoch": 2.4312428444218295, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.962018013000488, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8788547515869141, "num_tokens": 729260456.0, "step": 19112 }, { "epoch": 2.4313700547004196, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.008491516113281, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8624911904335022, "num_tokens": 729299654.0, "step": 19113 }, { "epoch": 2.4314972649790105, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.960198402404785, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8777024745941162, "num_tokens": 729341296.0, "step": 19114 }, { "epoch": 2.4316244752576006, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.07653522491455, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8776247501373291, "num_tokens": 729379632.0, "step": 19115 }, { "epoch": 2.431751685536191, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.991518020629883, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8772571086883545, "num_tokens": 729423202.0, "step": 19116 }, { "epoch": 2.4318788958147817, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.041454315185547, "learning_rate": 1e-06, "loss": 0.3331, "mean_token_accuracy": 0.8937534093856812, "num_tokens": 729457108.0, "step": 19117 }, { "epoch": 2.432006106093372, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.05062198638916, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.870586097240448, "num_tokens": 729494065.0, "step": 19118 }, { "epoch": 2.4321333163719627, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.04211711883545, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.878459095954895, "num_tokens": 729536621.0, "step": 19119 }, { "epoch": 2.4322605266505533, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.032185554504395, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.875213623046875, "num_tokens": 729574423.0, "step": 19120 }, { "epoch": 2.432387736929144, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.973814010620117, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8696278929710388, "num_tokens": 729614673.0, "step": 19121 }, { "epoch": 2.4325149472077343, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.01827335357666, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.876323938369751, "num_tokens": 729653049.0, "step": 19122 }, { "epoch": 2.432642157486325, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.053191184997559, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8747887015342712, "num_tokens": 729689667.0, "step": 19123 }, { "epoch": 2.4327693677649154, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.945625305175781, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8791183233261108, "num_tokens": 729734747.0, "step": 19124 }, { "epoch": 2.432896578043506, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.012673377990723, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8704737424850464, "num_tokens": 729770135.0, "step": 19125 }, { "epoch": 2.4330237883220964, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.98212718963623, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8732905387878418, "num_tokens": 729809203.0, "step": 19126 }, { "epoch": 2.433150998600687, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.071496963500977, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8804829120635986, "num_tokens": 729850164.0, "step": 19127 }, { "epoch": 2.4332782088792775, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.98464298248291, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8631752729415894, "num_tokens": 729884337.0, "step": 19128 }, { "epoch": 2.433405419157868, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.972203254699707, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8868918418884277, "num_tokens": 729924098.0, "step": 19129 }, { "epoch": 2.4335326294364585, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.958125114440918, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.8926506042480469, "num_tokens": 729960671.0, "step": 19130 }, { "epoch": 2.433659839715049, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.915942192077637, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8751309514045715, "num_tokens": 730000780.0, "step": 19131 }, { "epoch": 2.4337870499936396, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.989168167114258, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8728921413421631, "num_tokens": 730045198.0, "step": 19132 }, { "epoch": 2.43391426027223, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.021001815795898, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8678451180458069, "num_tokens": 730085010.0, "step": 19133 }, { "epoch": 2.4340414705508207, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.99283218383789, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8687835931777954, "num_tokens": 730121512.0, "step": 19134 }, { "epoch": 2.434168680829411, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.960763931274414, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8736507296562195, "num_tokens": 730164859.0, "step": 19135 }, { "epoch": 2.4342958911080017, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.004178047180176, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8747027516365051, "num_tokens": 730202273.0, "step": 19136 }, { "epoch": 2.4344231013865922, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.026674270629883, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.867398738861084, "num_tokens": 730248208.0, "step": 19137 }, { "epoch": 2.4345503116651823, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.999764442443848, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8834047317504883, "num_tokens": 730291084.0, "step": 19138 }, { "epoch": 2.4346775219437733, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.000418663024902, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8780481815338135, "num_tokens": 730334504.0, "step": 19139 }, { "epoch": 2.4348047322223634, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.013848304748535, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.878214955329895, "num_tokens": 730373356.0, "step": 19140 }, { "epoch": 2.434931942500954, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.008878707885742, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8620644211769104, "num_tokens": 730416229.0, "step": 19141 }, { "epoch": 2.4350591527795444, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.005877494812012, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8862742185592651, "num_tokens": 730458922.0, "step": 19142 }, { "epoch": 2.435186363058135, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.064078330993652, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8732329607009888, "num_tokens": 730494583.0, "step": 19143 }, { "epoch": 2.4353135733367255, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.994011878967285, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8771003484725952, "num_tokens": 730529526.0, "step": 19144 }, { "epoch": 2.435440783615316, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.005244255065918, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.877930223941803, "num_tokens": 730558584.0, "step": 19145 }, { "epoch": 2.4355679938939065, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.032637596130371, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8753517866134644, "num_tokens": 730598677.0, "step": 19146 }, { "epoch": 2.435695204172497, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.000332832336426, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8839581608772278, "num_tokens": 730636985.0, "step": 19147 }, { "epoch": 2.4358224144510876, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.059337615966797, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8572803735733032, "num_tokens": 730675409.0, "step": 19148 }, { "epoch": 2.435949624729678, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.017759323120117, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8637480735778809, "num_tokens": 730714739.0, "step": 19149 }, { "epoch": 2.4360768350082687, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.99038028717041, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8739360570907593, "num_tokens": 730757059.0, "step": 19150 }, { "epoch": 2.436204045286859, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.002583503723145, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8742530345916748, "num_tokens": 730793469.0, "step": 19151 }, { "epoch": 2.4363312555654497, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.025981903076172, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8795125484466553, "num_tokens": 730829336.0, "step": 19152 }, { "epoch": 2.4364584658440402, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.952493667602539, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8725194931030273, "num_tokens": 730869304.0, "step": 19153 }, { "epoch": 2.4365856761226308, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.021686553955078, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8622296452522278, "num_tokens": 730905315.0, "step": 19154 }, { "epoch": 2.4367128864012213, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.99687671661377, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.865138053894043, "num_tokens": 730947346.0, "step": 19155 }, { "epoch": 2.436840096679812, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.989730834960938, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8781310319900513, "num_tokens": 730976907.0, "step": 19156 }, { "epoch": 2.4369673069584024, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.934361457824707, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8709558248519897, "num_tokens": 731015940.0, "step": 19157 }, { "epoch": 2.437094517236993, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.00580883026123, "learning_rate": 1e-06, "loss": 0.3349, "mean_token_accuracy": 0.8902190327644348, "num_tokens": 731051345.0, "step": 19158 }, { "epoch": 2.4372217275155834, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.019906997680664, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.8441617488861084, "num_tokens": 731089642.0, "step": 19159 }, { "epoch": 2.437348937794174, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.975116729736328, "learning_rate": 1e-06, "loss": 0.347, "mean_token_accuracy": 0.8911532759666443, "num_tokens": 731133196.0, "step": 19160 }, { "epoch": 2.4374761480727645, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.956950187683105, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8626759052276611, "num_tokens": 731175510.0, "step": 19161 }, { "epoch": 2.437603358351355, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.012585639953613, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8822893500328064, "num_tokens": 731217526.0, "step": 19162 }, { "epoch": 2.437730568629945, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.942157745361328, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8625980615615845, "num_tokens": 731259525.0, "step": 19163 }, { "epoch": 2.437857778908536, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.000813484191895, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.877819299697876, "num_tokens": 731301480.0, "step": 19164 }, { "epoch": 2.437984989187126, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.948423385620117, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8831157088279724, "num_tokens": 731347385.0, "step": 19165 }, { "epoch": 2.4381121994657167, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.077884674072266, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8789799213409424, "num_tokens": 731383666.0, "step": 19166 }, { "epoch": 2.438239409744307, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.023736953735352, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8748685121536255, "num_tokens": 731420297.0, "step": 19167 }, { "epoch": 2.4383666200228977, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.034977912902832, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8709553480148315, "num_tokens": 731455083.0, "step": 19168 }, { "epoch": 2.4384938303014883, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.057613372802734, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8821215629577637, "num_tokens": 731482555.0, "step": 19169 }, { "epoch": 2.438621040580079, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.017760276794434, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8861187696456909, "num_tokens": 731519459.0, "step": 19170 }, { "epoch": 2.4387482508586693, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.926443099975586, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8794287443161011, "num_tokens": 731554577.0, "step": 19171 }, { "epoch": 2.43887546113726, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.012382507324219, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8659145832061768, "num_tokens": 731594993.0, "step": 19172 }, { "epoch": 2.4390026714158504, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.038816452026367, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8786540031433105, "num_tokens": 731629182.0, "step": 19173 }, { "epoch": 2.439129881694441, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.99485969543457, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8733149170875549, "num_tokens": 731667661.0, "step": 19174 }, { "epoch": 2.4392570919730314, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.024850845336914, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8748812675476074, "num_tokens": 731704911.0, "step": 19175 }, { "epoch": 2.439384302251622, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.026022911071777, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8586762547492981, "num_tokens": 731744990.0, "step": 19176 }, { "epoch": 2.4395115125302125, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.000744819641113, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8753525614738464, "num_tokens": 731783335.0, "step": 19177 }, { "epoch": 2.439638722808803, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.988784790039062, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8769482970237732, "num_tokens": 731819206.0, "step": 19178 }, { "epoch": 2.4397659330873935, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.01738452911377, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8723358511924744, "num_tokens": 731855291.0, "step": 19179 }, { "epoch": 2.439893143365984, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.93398380279541, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8785311579704285, "num_tokens": 731899392.0, "step": 19180 }, { "epoch": 2.4400203536445746, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.015398025512695, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8736611604690552, "num_tokens": 731937502.0, "step": 19181 }, { "epoch": 2.440147563923165, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.872220993041992, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8775396347045898, "num_tokens": 731974946.0, "step": 19182 }, { "epoch": 2.4402747742017556, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.99995231628418, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8630197644233704, "num_tokens": 732018732.0, "step": 19183 }, { "epoch": 2.440401984480346, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.982610702514648, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8736462593078613, "num_tokens": 732052036.0, "step": 19184 }, { "epoch": 2.4405291947589367, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.09395694732666, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8695718050003052, "num_tokens": 732090944.0, "step": 19185 }, { "epoch": 2.440656405037527, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.990191459655762, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8771889209747314, "num_tokens": 732131406.0, "step": 19186 }, { "epoch": 2.4407836153161178, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.045720100402832, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8734523057937622, "num_tokens": 732171701.0, "step": 19187 }, { "epoch": 2.440910825594708, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.0267333984375, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8828896284103394, "num_tokens": 732209658.0, "step": 19188 }, { "epoch": 2.4410380358732984, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.075658798217773, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8626950979232788, "num_tokens": 732251395.0, "step": 19189 }, { "epoch": 2.441165246151889, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.0372953414917, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8642690777778625, "num_tokens": 732289306.0, "step": 19190 }, { "epoch": 2.4412924564304794, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.998833656311035, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.859886646270752, "num_tokens": 732337916.0, "step": 19191 }, { "epoch": 2.44141966670907, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.968564987182617, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8604695796966553, "num_tokens": 732380358.0, "step": 19192 }, { "epoch": 2.4415468769876605, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.013556480407715, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8843882083892822, "num_tokens": 732416534.0, "step": 19193 }, { "epoch": 2.441674087266251, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.053447723388672, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8542174696922302, "num_tokens": 732458927.0, "step": 19194 }, { "epoch": 2.4418012975448415, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.047934532165527, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8662753105163574, "num_tokens": 732501261.0, "step": 19195 }, { "epoch": 2.441928507823432, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.966911315917969, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8816552758216858, "num_tokens": 732541162.0, "step": 19196 }, { "epoch": 2.4420557181020226, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.1094970703125, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8711892366409302, "num_tokens": 732578237.0, "step": 19197 }, { "epoch": 2.442182928380613, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.982977867126465, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8644005656242371, "num_tokens": 732618624.0, "step": 19198 }, { "epoch": 2.4423101386592037, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.056694984436035, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8698622584342957, "num_tokens": 732661747.0, "step": 19199 }, { "epoch": 2.442437348937794, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.023104667663574, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8536889553070068, "num_tokens": 732703511.0, "step": 19200 }, { "epoch": 2.4425645592163847, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.03289794921875, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8542211651802063, "num_tokens": 732746749.0, "step": 19201 }, { "epoch": 2.4426917694949752, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.027488708496094, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8761661052703857, "num_tokens": 732783184.0, "step": 19202 }, { "epoch": 2.4428189797735658, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.924498558044434, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8684437870979309, "num_tokens": 732818664.0, "step": 19203 }, { "epoch": 2.4429461900521563, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.09346866607666, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8635889291763306, "num_tokens": 732854286.0, "step": 19204 }, { "epoch": 2.443073400330747, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.959349632263184, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.880922257900238, "num_tokens": 732893032.0, "step": 19205 }, { "epoch": 2.4432006106093374, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.05871868133545, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8737844228744507, "num_tokens": 732929942.0, "step": 19206 }, { "epoch": 2.443327820887928, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.959813117980957, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8697904348373413, "num_tokens": 732968218.0, "step": 19207 }, { "epoch": 2.4434550311665184, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.074405670166016, "learning_rate": 1e-06, "loss": 0.3235, "mean_token_accuracy": 0.8968486785888672, "num_tokens": 733001464.0, "step": 19208 }, { "epoch": 2.443582241445109, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.994184494018555, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8649628162384033, "num_tokens": 733036311.0, "step": 19209 }, { "epoch": 2.4437094517236995, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.021162033081055, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.879816472530365, "num_tokens": 733070487.0, "step": 19210 }, { "epoch": 2.4438366620022896, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.996992111206055, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.87176913022995, "num_tokens": 733110357.0, "step": 19211 }, { "epoch": 2.4439638722808805, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.025471687316895, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8667925596237183, "num_tokens": 733149070.0, "step": 19212 }, { "epoch": 2.4440910825594706, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.991650581359863, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8894802331924438, "num_tokens": 733180848.0, "step": 19213 }, { "epoch": 2.444218292838061, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.012752532958984, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8752930760383606, "num_tokens": 733219204.0, "step": 19214 }, { "epoch": 2.4443455031166517, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.027934074401855, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8633363246917725, "num_tokens": 733255237.0, "step": 19215 }, { "epoch": 2.444472713395242, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.979559898376465, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8848053216934204, "num_tokens": 733293243.0, "step": 19216 }, { "epoch": 2.4445999236738327, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.000358581542969, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8770644664764404, "num_tokens": 733333530.0, "step": 19217 }, { "epoch": 2.4447271339524232, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.111241340637207, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8558803796768188, "num_tokens": 733373790.0, "step": 19218 }, { "epoch": 2.4448543442310138, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.025827407836914, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8764876127243042, "num_tokens": 733416037.0, "step": 19219 }, { "epoch": 2.4449815545096043, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.066286087036133, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8702635765075684, "num_tokens": 733453014.0, "step": 19220 }, { "epoch": 2.445108764788195, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.040067672729492, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8588907122612, "num_tokens": 733491197.0, "step": 19221 }, { "epoch": 2.4452359750667854, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.092229843139648, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8862189054489136, "num_tokens": 733531142.0, "step": 19222 }, { "epoch": 2.445363185345376, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.09886646270752, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8681999444961548, "num_tokens": 733573916.0, "step": 19223 }, { "epoch": 2.4454903956239664, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.101546287536621, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8678651452064514, "num_tokens": 733613204.0, "step": 19224 }, { "epoch": 2.445617605902557, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.084037780761719, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8691615462303162, "num_tokens": 733653618.0, "step": 19225 }, { "epoch": 2.4457448161811475, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 14.060600280761719, "learning_rate": 1e-06, "loss": 0.3519, "mean_token_accuracy": 0.8905457258224487, "num_tokens": 733687795.0, "step": 19226 }, { "epoch": 2.445872026459738, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.09240436553955, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8734050989151001, "num_tokens": 733723574.0, "step": 19227 }, { "epoch": 2.4459992367383285, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.029353141784668, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8803369402885437, "num_tokens": 733762051.0, "step": 19228 }, { "epoch": 2.446126447016919, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.997773170471191, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.8898577690124512, "num_tokens": 733800640.0, "step": 19229 }, { "epoch": 2.4462536572955096, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.980093955993652, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8632367849349976, "num_tokens": 733841321.0, "step": 19230 }, { "epoch": 2.4463808675741, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.034311294555664, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8767350912094116, "num_tokens": 733874104.0, "step": 19231 }, { "epoch": 2.4465080778526906, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.981229782104492, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8781865835189819, "num_tokens": 733913127.0, "step": 19232 }, { "epoch": 2.446635288131281, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.02182674407959, "learning_rate": 1e-06, "loss": 0.345, "mean_token_accuracy": 0.8905974626541138, "num_tokens": 733951900.0, "step": 19233 }, { "epoch": 2.4467624984098717, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.043316841125488, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.876596212387085, "num_tokens": 733992866.0, "step": 19234 }, { "epoch": 2.4468897086884622, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.021839141845703, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8689366579055786, "num_tokens": 734029895.0, "step": 19235 }, { "epoch": 2.4470169189670523, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.019268989562988, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8908647298812866, "num_tokens": 734066954.0, "step": 19236 }, { "epoch": 2.4471441292456433, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.065004348754883, "learning_rate": 1e-06, "loss": 0.3328, "mean_token_accuracy": 0.8953306674957275, "num_tokens": 734102412.0, "step": 19237 }, { "epoch": 2.4472713395242334, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.123961448669434, "learning_rate": 1e-06, "loss": 0.3443, "mean_token_accuracy": 0.8900331258773804, "num_tokens": 734141400.0, "step": 19238 }, { "epoch": 2.447398549802824, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.095813751220703, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.854634165763855, "num_tokens": 734176628.0, "step": 19239 }, { "epoch": 2.4475257600814144, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.043805122375488, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8796095848083496, "num_tokens": 734219756.0, "step": 19240 }, { "epoch": 2.447652970360005, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.067814826965332, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8736101388931274, "num_tokens": 734250509.0, "step": 19241 }, { "epoch": 2.4477801806385955, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.06314468383789, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8812774419784546, "num_tokens": 734287887.0, "step": 19242 }, { "epoch": 2.447907390917186, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.062368392944336, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8780596256256104, "num_tokens": 734330446.0, "step": 19243 }, { "epoch": 2.4480346011957765, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.963741302490234, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8791341781616211, "num_tokens": 734368427.0, "step": 19244 }, { "epoch": 2.448161811474367, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.02711296081543, "learning_rate": 1e-06, "loss": 0.3362, "mean_token_accuracy": 0.8927149772644043, "num_tokens": 734405103.0, "step": 19245 }, { "epoch": 2.4482890217529576, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.082853317260742, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8711287975311279, "num_tokens": 734448069.0, "step": 19246 }, { "epoch": 2.448416232031548, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.055273056030273, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8610531091690063, "num_tokens": 734486016.0, "step": 19247 }, { "epoch": 2.4485434423101387, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.022394180297852, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8632071018218994, "num_tokens": 734526943.0, "step": 19248 }, { "epoch": 2.448670652588729, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.00711727142334, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8879859447479248, "num_tokens": 734566137.0, "step": 19249 }, { "epoch": 2.4487978628673197, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.034940719604492, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8723300695419312, "num_tokens": 734602366.0, "step": 19250 }, { "epoch": 2.4489250731459102, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.03304386138916, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8821997046470642, "num_tokens": 734644251.0, "step": 19251 }, { "epoch": 2.4490522834245008, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.038202285766602, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8702488541603088, "num_tokens": 734682938.0, "step": 19252 }, { "epoch": 2.4491794937030913, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.044933319091797, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8754458427429199, "num_tokens": 734722975.0, "step": 19253 }, { "epoch": 2.449306703981682, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.025795936584473, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8774129152297974, "num_tokens": 734761361.0, "step": 19254 }, { "epoch": 2.4494339142602723, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.037487030029297, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.873021125793457, "num_tokens": 734800597.0, "step": 19255 }, { "epoch": 2.449561124538863, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.058000564575195, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.86320561170578, "num_tokens": 734840575.0, "step": 19256 }, { "epoch": 2.4496883348174534, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.02681827545166, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8659256100654602, "num_tokens": 734876943.0, "step": 19257 }, { "epoch": 2.449815545096044, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.0338716506958, "learning_rate": 1e-06, "loss": 0.348, "mean_token_accuracy": 0.8893131613731384, "num_tokens": 734916436.0, "step": 19258 }, { "epoch": 2.4499427553746345, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.043992042541504, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8730664849281311, "num_tokens": 734956336.0, "step": 19259 }, { "epoch": 2.450069965653225, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.06295394897461, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8728737831115723, "num_tokens": 734997132.0, "step": 19260 }, { "epoch": 2.450197175931815, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.094819068908691, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8628268241882324, "num_tokens": 735036192.0, "step": 19261 }, { "epoch": 2.450324386210406, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.959203720092773, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8874509334564209, "num_tokens": 735079368.0, "step": 19262 }, { "epoch": 2.450451596488996, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.115865707397461, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8798239827156067, "num_tokens": 735109714.0, "step": 19263 }, { "epoch": 2.4505788067675867, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.051204681396484, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.871559739112854, "num_tokens": 735152712.0, "step": 19264 }, { "epoch": 2.450706017046177, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.006229400634766, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8692200183868408, "num_tokens": 735193313.0, "step": 19265 }, { "epoch": 2.4508332273247677, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.985307693481445, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8883281946182251, "num_tokens": 735235413.0, "step": 19266 }, { "epoch": 2.4509604376033582, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.056812286376953, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8754384517669678, "num_tokens": 735270809.0, "step": 19267 }, { "epoch": 2.4510876478819488, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.067560195922852, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8679460883140564, "num_tokens": 735309790.0, "step": 19268 }, { "epoch": 2.4512148581605393, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.05200481414795, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8751468658447266, "num_tokens": 735341084.0, "step": 19269 }, { "epoch": 2.45134206843913, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.040837287902832, "learning_rate": 1e-06, "loss": 0.3452, "mean_token_accuracy": 0.8901399374008179, "num_tokens": 735376964.0, "step": 19270 }, { "epoch": 2.4514692787177204, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.060713768005371, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8838492035865784, "num_tokens": 735413296.0, "step": 19271 }, { "epoch": 2.451596488996311, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.034643173217773, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8786998987197876, "num_tokens": 735454992.0, "step": 19272 }, { "epoch": 2.4517236992749014, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.026690483093262, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8763175010681152, "num_tokens": 735493315.0, "step": 19273 }, { "epoch": 2.451850909553492, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.051438331604004, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8823411464691162, "num_tokens": 735532666.0, "step": 19274 }, { "epoch": 2.4519781198320825, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.062932968139648, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8729528188705444, "num_tokens": 735571987.0, "step": 19275 }, { "epoch": 2.452105330110673, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.04804515838623, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8722019195556641, "num_tokens": 735609409.0, "step": 19276 }, { "epoch": 2.4522325403892635, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.081221580505371, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8486889600753784, "num_tokens": 735647299.0, "step": 19277 }, { "epoch": 2.452359750667854, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.085955619812012, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8818720579147339, "num_tokens": 735677757.0, "step": 19278 }, { "epoch": 2.4524869609464446, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.054460525512695, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8801824450492859, "num_tokens": 735723924.0, "step": 19279 }, { "epoch": 2.452614171225035, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.139167785644531, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8594343662261963, "num_tokens": 735758105.0, "step": 19280 }, { "epoch": 2.4527413815036256, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.085166931152344, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8709561824798584, "num_tokens": 735797603.0, "step": 19281 }, { "epoch": 2.452868591782216, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.058870315551758, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.8480391502380371, "num_tokens": 735838958.0, "step": 19282 }, { "epoch": 2.4529958020608067, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.148971557617188, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8667958378791809, "num_tokens": 735878894.0, "step": 19283 }, { "epoch": 2.453123012339397, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.029982566833496, "learning_rate": 1e-06, "loss": 0.3073, "mean_token_accuracy": 0.9037951827049255, "num_tokens": 735913222.0, "step": 19284 }, { "epoch": 2.4532502226179878, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.12010383605957, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8818157911300659, "num_tokens": 735951884.0, "step": 19285 }, { "epoch": 2.453377432896578, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.9849214553833, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8703950047492981, "num_tokens": 735998955.0, "step": 19286 }, { "epoch": 2.4535046431751684, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.009929656982422, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8733069896697998, "num_tokens": 736039631.0, "step": 19287 }, { "epoch": 2.453631853453759, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.076265335083008, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8697505593299866, "num_tokens": 736085100.0, "step": 19288 }, { "epoch": 2.4537590637323494, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.054102897644043, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8891924619674683, "num_tokens": 736119680.0, "step": 19289 }, { "epoch": 2.45388627401094, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.062713623046875, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8866260647773743, "num_tokens": 736159769.0, "step": 19290 }, { "epoch": 2.4540134842895305, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.077720642089844, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8796286582946777, "num_tokens": 736199566.0, "step": 19291 }, { "epoch": 2.454140694568121, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.039131164550781, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8785505294799805, "num_tokens": 736239238.0, "step": 19292 }, { "epoch": 2.4542679048467115, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.060890197753906, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8777651190757751, "num_tokens": 736274576.0, "step": 19293 }, { "epoch": 2.454395115125302, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.05539608001709, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.869184672832489, "num_tokens": 736307509.0, "step": 19294 }, { "epoch": 2.4545223254038926, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.084171295166016, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8705506920814514, "num_tokens": 736348313.0, "step": 19295 }, { "epoch": 2.454649535682483, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.068388938903809, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8706570863723755, "num_tokens": 736380716.0, "step": 19296 }, { "epoch": 2.4547767459610736, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.016436576843262, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8534190654754639, "num_tokens": 736422669.0, "step": 19297 }, { "epoch": 2.454903956239664, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.111489295959473, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8816887140274048, "num_tokens": 736452929.0, "step": 19298 }, { "epoch": 2.4550311665182547, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.954136848449707, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8760142922401428, "num_tokens": 736486755.0, "step": 19299 }, { "epoch": 2.4551583767968452, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.016660690307617, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8832433223724365, "num_tokens": 736521542.0, "step": 19300 }, { "epoch": 2.4552855870754358, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.039194107055664, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8761576414108276, "num_tokens": 736557312.0, "step": 19301 }, { "epoch": 2.4554127973540263, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.028480529785156, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8699531555175781, "num_tokens": 736596469.0, "step": 19302 }, { "epoch": 2.455540007632617, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.092267036437988, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8598535060882568, "num_tokens": 736636078.0, "step": 19303 }, { "epoch": 2.4556672179112073, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.02082347869873, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8867178559303284, "num_tokens": 736674117.0, "step": 19304 }, { "epoch": 2.455794428189798, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.110376358032227, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8714572787284851, "num_tokens": 736709116.0, "step": 19305 }, { "epoch": 2.4559216384683884, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.972053527832031, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.866377055644989, "num_tokens": 736750383.0, "step": 19306 }, { "epoch": 2.456048848746979, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.14015007019043, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8636873960494995, "num_tokens": 736790836.0, "step": 19307 }, { "epoch": 2.4561760590255695, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.99414348602295, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8727878928184509, "num_tokens": 736826085.0, "step": 19308 }, { "epoch": 2.4563032693041595, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.077789306640625, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8742015957832336, "num_tokens": 736860599.0, "step": 19309 }, { "epoch": 2.4564304795827505, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.054876327514648, "learning_rate": 1e-06, "loss": 0.335, "mean_token_accuracy": 0.8914228677749634, "num_tokens": 736893286.0, "step": 19310 }, { "epoch": 2.4565576898613406, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.039155006408691, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8834924101829529, "num_tokens": 736931597.0, "step": 19311 }, { "epoch": 2.456684900139931, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.018338203430176, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8819268941879272, "num_tokens": 736972183.0, "step": 19312 }, { "epoch": 2.4568121104185217, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.028183937072754, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8809099197387695, "num_tokens": 737004184.0, "step": 19313 }, { "epoch": 2.456939320697112, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.054082870483398, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8741583228111267, "num_tokens": 737042644.0, "step": 19314 }, { "epoch": 2.4570665309757027, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.1522798538208, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8651520013809204, "num_tokens": 737074830.0, "step": 19315 }, { "epoch": 2.4571937412542932, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.134211540222168, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8817828297615051, "num_tokens": 737107756.0, "step": 19316 }, { "epoch": 2.4573209515328838, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.049060821533203, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8835182189941406, "num_tokens": 737143457.0, "step": 19317 }, { "epoch": 2.4574481618114743, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.06343936920166, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.870177149772644, "num_tokens": 737177005.0, "step": 19318 }, { "epoch": 2.457575372090065, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.06495189666748, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.865602970123291, "num_tokens": 737218265.0, "step": 19319 }, { "epoch": 2.4577025823686554, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.0982027053833, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.864986002445221, "num_tokens": 737259495.0, "step": 19320 }, { "epoch": 2.457829792647246, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.988081932067871, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8664277791976929, "num_tokens": 737300073.0, "step": 19321 }, { "epoch": 2.4579570029258364, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.142441749572754, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8759475946426392, "num_tokens": 737334716.0, "step": 19322 }, { "epoch": 2.458084213204427, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.101350784301758, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8792486190795898, "num_tokens": 737377086.0, "step": 19323 }, { "epoch": 2.4582114234830175, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.056562423706055, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.864851713180542, "num_tokens": 737418259.0, "step": 19324 }, { "epoch": 2.458338633761608, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.111356735229492, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8654552102088928, "num_tokens": 737453150.0, "step": 19325 }, { "epoch": 2.4584658440401985, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.016599655151367, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8893024921417236, "num_tokens": 737495552.0, "step": 19326 }, { "epoch": 2.458593054318789, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.13744068145752, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.880695641040802, "num_tokens": 737537518.0, "step": 19327 }, { "epoch": 2.4587202645973796, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.032806396484375, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.870579719543457, "num_tokens": 737577472.0, "step": 19328 }, { "epoch": 2.45884747487597, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.085806846618652, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8648553490638733, "num_tokens": 737617250.0, "step": 19329 }, { "epoch": 2.4589746851545606, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.092947959899902, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8777408003807068, "num_tokens": 737655544.0, "step": 19330 }, { "epoch": 2.459101895433151, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.018501281738281, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8779067993164062, "num_tokens": 737688644.0, "step": 19331 }, { "epoch": 2.4592291057117417, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.020968437194824, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8787494897842407, "num_tokens": 737730264.0, "step": 19332 }, { "epoch": 2.459356315990332, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.133098602294922, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8688337206840515, "num_tokens": 737763578.0, "step": 19333 }, { "epoch": 2.4594835262689223, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.079315185546875, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.875849723815918, "num_tokens": 737804062.0, "step": 19334 }, { "epoch": 2.4596107365475133, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.043306350708008, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8884599208831787, "num_tokens": 737845608.0, "step": 19335 }, { "epoch": 2.4597379468261034, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.117474555969238, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8570472002029419, "num_tokens": 737881366.0, "step": 19336 }, { "epoch": 2.459865157104694, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.047977447509766, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8670126795768738, "num_tokens": 737922595.0, "step": 19337 }, { "epoch": 2.4599923673832844, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.079488754272461, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8813749551773071, "num_tokens": 737956368.0, "step": 19338 }, { "epoch": 2.460119577661875, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.993741989135742, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8788156509399414, "num_tokens": 738000167.0, "step": 19339 }, { "epoch": 2.4602467879404655, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.070442199707031, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8673056960105896, "num_tokens": 738040850.0, "step": 19340 }, { "epoch": 2.460373998219056, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.054046630859375, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8728339076042175, "num_tokens": 738080370.0, "step": 19341 }, { "epoch": 2.4605012084976465, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.057190895080566, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8797980546951294, "num_tokens": 738116751.0, "step": 19342 }, { "epoch": 2.460628418776237, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.06640911102295, "learning_rate": 1e-06, "loss": 0.3244, "mean_token_accuracy": 0.8916705846786499, "num_tokens": 738154597.0, "step": 19343 }, { "epoch": 2.4607556290548276, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.046977043151855, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8813004493713379, "num_tokens": 738186484.0, "step": 19344 }, { "epoch": 2.460882839333418, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.103446960449219, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8787519335746765, "num_tokens": 738217722.0, "step": 19345 }, { "epoch": 2.4610100496120086, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.064118385314941, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8879089951515198, "num_tokens": 738255304.0, "step": 19346 }, { "epoch": 2.461137259890599, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.026869773864746, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8699302077293396, "num_tokens": 738292162.0, "step": 19347 }, { "epoch": 2.4612644701691897, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.103204727172852, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.886674702167511, "num_tokens": 738332851.0, "step": 19348 }, { "epoch": 2.4613916804477802, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.068428993225098, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8845803141593933, "num_tokens": 738375120.0, "step": 19349 }, { "epoch": 2.4615188907263708, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.051481246948242, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8658651113510132, "num_tokens": 738413857.0, "step": 19350 }, { "epoch": 2.4616461010049613, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.175983428955078, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8691372275352478, "num_tokens": 738450097.0, "step": 19351 }, { "epoch": 2.461773311283552, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.011065483093262, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8565660119056702, "num_tokens": 738483784.0, "step": 19352 }, { "epoch": 2.4619005215621423, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.110071182250977, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8562083840370178, "num_tokens": 738526820.0, "step": 19353 }, { "epoch": 2.462027731840733, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.993494987487793, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8605069518089294, "num_tokens": 738569049.0, "step": 19354 }, { "epoch": 2.4621549421193234, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.067358016967773, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.871303915977478, "num_tokens": 738608320.0, "step": 19355 }, { "epoch": 2.462282152397914, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.044673919677734, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8892949223518372, "num_tokens": 738645566.0, "step": 19356 }, { "epoch": 2.4624093626765045, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.058119773864746, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8519565463066101, "num_tokens": 738686184.0, "step": 19357 }, { "epoch": 2.462536572955095, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.022948265075684, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8615841269493103, "num_tokens": 738728820.0, "step": 19358 }, { "epoch": 2.462663783233685, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.096442222595215, "learning_rate": 1e-06, "loss": 0.4859, "mean_token_accuracy": 0.8456686735153198, "num_tokens": 738760180.0, "step": 19359 }, { "epoch": 2.462790993512276, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.093851089477539, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8822369575500488, "num_tokens": 738795501.0, "step": 19360 }, { "epoch": 2.462918203790866, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.988018035888672, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8654620051383972, "num_tokens": 738831027.0, "step": 19361 }, { "epoch": 2.4630454140694567, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.099302291870117, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8574764132499695, "num_tokens": 738870383.0, "step": 19362 }, { "epoch": 2.463172624348047, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.062270164489746, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.878529965877533, "num_tokens": 738908231.0, "step": 19363 }, { "epoch": 2.4632998346266377, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.031930923461914, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8655171394348145, "num_tokens": 738950927.0, "step": 19364 }, { "epoch": 2.4634270449052282, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.990355491638184, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8857104778289795, "num_tokens": 738989857.0, "step": 19365 }, { "epoch": 2.4635542551838188, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.088883399963379, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.887597918510437, "num_tokens": 739021281.0, "step": 19366 }, { "epoch": 2.4636814654624093, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.52635669708252, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.881382942199707, "num_tokens": 739063475.0, "step": 19367 }, { "epoch": 2.463808675741, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.986030578613281, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8695564270019531, "num_tokens": 739107083.0, "step": 19368 }, { "epoch": 2.4639358860195903, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.035839080810547, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8536174893379211, "num_tokens": 739152002.0, "step": 19369 }, { "epoch": 2.464063096298181, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.096410751342773, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8721344470977783, "num_tokens": 739191781.0, "step": 19370 }, { "epoch": 2.4641903065767714, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.055225372314453, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.8941373825073242, "num_tokens": 739230374.0, "step": 19371 }, { "epoch": 2.464317516855362, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.06332778930664, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8712157607078552, "num_tokens": 739267310.0, "step": 19372 }, { "epoch": 2.4644447271339525, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.028885841369629, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8767343759536743, "num_tokens": 739305074.0, "step": 19373 }, { "epoch": 2.464571937412543, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.120686531066895, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8665828704833984, "num_tokens": 739343028.0, "step": 19374 }, { "epoch": 2.4646991476911335, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.121728897094727, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8731604218482971, "num_tokens": 739384364.0, "step": 19375 }, { "epoch": 2.464826357969724, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.201298713684082, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8899590969085693, "num_tokens": 739423564.0, "step": 19376 }, { "epoch": 2.4649535682483146, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.125638961791992, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.882840096950531, "num_tokens": 739457609.0, "step": 19377 }, { "epoch": 2.465080778526905, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.096165657043457, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8744561076164246, "num_tokens": 739497204.0, "step": 19378 }, { "epoch": 2.4652079888054956, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.114042282104492, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8688738346099854, "num_tokens": 739543688.0, "step": 19379 }, { "epoch": 2.465335199084086, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.10725212097168, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8859956860542297, "num_tokens": 739578221.0, "step": 19380 }, { "epoch": 2.4654624093626767, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.094694137573242, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8799724578857422, "num_tokens": 739615892.0, "step": 19381 }, { "epoch": 2.4655896196412668, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.070273399353027, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8601632118225098, "num_tokens": 739657697.0, "step": 19382 }, { "epoch": 2.4657168299198577, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.103292465209961, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8830215930938721, "num_tokens": 739694173.0, "step": 19383 }, { "epoch": 2.465844040198448, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.052840232849121, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8716781139373779, "num_tokens": 739734111.0, "step": 19384 }, { "epoch": 2.4659712504770384, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.147940635681152, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8754947185516357, "num_tokens": 739771409.0, "step": 19385 }, { "epoch": 2.466098460755629, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.102402687072754, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8609189391136169, "num_tokens": 739807185.0, "step": 19386 }, { "epoch": 2.4662256710342194, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.095383644104004, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8872823715209961, "num_tokens": 739837750.0, "step": 19387 }, { "epoch": 2.46635288131281, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.078322410583496, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.872733473777771, "num_tokens": 739879516.0, "step": 19388 }, { "epoch": 2.4664800915914005, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.988454818725586, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8807485103607178, "num_tokens": 739918688.0, "step": 19389 }, { "epoch": 2.466607301869991, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.073505401611328, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8779693841934204, "num_tokens": 739958978.0, "step": 19390 }, { "epoch": 2.4667345121485815, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.03575611114502, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8792253732681274, "num_tokens": 740002428.0, "step": 19391 }, { "epoch": 2.466861722427172, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.080587387084961, "learning_rate": 1e-06, "loss": 0.3416, "mean_token_accuracy": 0.8945554494857788, "num_tokens": 740045856.0, "step": 19392 }, { "epoch": 2.4669889327057626, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.089109420776367, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8722081184387207, "num_tokens": 740084512.0, "step": 19393 }, { "epoch": 2.467116142984353, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.974692344665527, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8850530385971069, "num_tokens": 740129624.0, "step": 19394 }, { "epoch": 2.4672433532629436, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.126025199890137, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8583476543426514, "num_tokens": 740168368.0, "step": 19395 }, { "epoch": 2.467370563541534, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.12458324432373, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8679002523422241, "num_tokens": 740206670.0, "step": 19396 }, { "epoch": 2.4674977738201247, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.092631340026855, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8530803918838501, "num_tokens": 740246422.0, "step": 19397 }, { "epoch": 2.4676249840987152, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.112528800964355, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8719422817230225, "num_tokens": 740284450.0, "step": 19398 }, { "epoch": 2.4677521943773058, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.073892593383789, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8516439199447632, "num_tokens": 740319062.0, "step": 19399 }, { "epoch": 2.4678794046558963, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.002828598022461, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8752321004867554, "num_tokens": 740353996.0, "step": 19400 }, { "epoch": 2.468006614934487, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.110503196716309, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8733322620391846, "num_tokens": 740394694.0, "step": 19401 }, { "epoch": 2.4681338252130773, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.012210845947266, "learning_rate": 1e-06, "loss": 0.3395, "mean_token_accuracy": 0.8940869569778442, "num_tokens": 740436567.0, "step": 19402 }, { "epoch": 2.468261035491668, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.163895606994629, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8671696186065674, "num_tokens": 740476458.0, "step": 19403 }, { "epoch": 2.4683882457702584, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.017620086669922, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8823792934417725, "num_tokens": 740514703.0, "step": 19404 }, { "epoch": 2.468515456048849, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 14.05225658416748, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8846819996833801, "num_tokens": 740552328.0, "step": 19405 }, { "epoch": 2.4686426663274394, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.090587615966797, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8569658994674683, "num_tokens": 740593141.0, "step": 19406 }, { "epoch": 2.4687698766060295, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.975020408630371, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8847554326057434, "num_tokens": 740634097.0, "step": 19407 }, { "epoch": 2.4688970868846205, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.137890815734863, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.879571259021759, "num_tokens": 740672917.0, "step": 19408 }, { "epoch": 2.4690242971632106, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.994561195373535, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8694776296615601, "num_tokens": 740709351.0, "step": 19409 }, { "epoch": 2.469151507441801, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.084049224853516, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8643954992294312, "num_tokens": 740744049.0, "step": 19410 }, { "epoch": 2.4692787177203916, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.993939399719238, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8772220611572266, "num_tokens": 740784765.0, "step": 19411 }, { "epoch": 2.469405927998982, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.109670639038086, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8680299520492554, "num_tokens": 740815859.0, "step": 19412 }, { "epoch": 2.4695331382775727, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.12136173248291, "learning_rate": 1e-06, "loss": 0.3358, "mean_token_accuracy": 0.8916551470756531, "num_tokens": 740854300.0, "step": 19413 }, { "epoch": 2.4696603485561632, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.065629959106445, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.858043372631073, "num_tokens": 740891836.0, "step": 19414 }, { "epoch": 2.4697875588347538, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.036369323730469, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8827167749404907, "num_tokens": 740927434.0, "step": 19415 }, { "epoch": 2.4699147691133443, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.016166687011719, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8768934607505798, "num_tokens": 740973707.0, "step": 19416 }, { "epoch": 2.470041979391935, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.108187675476074, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8814849257469177, "num_tokens": 741015953.0, "step": 19417 }, { "epoch": 2.4701691896705253, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.125743865966797, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.88158118724823, "num_tokens": 741049634.0, "step": 19418 }, { "epoch": 2.470296399949116, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.03706169128418, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8844764828681946, "num_tokens": 741084267.0, "step": 19419 }, { "epoch": 2.4704236102277064, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.102729797363281, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8805533647537231, "num_tokens": 741122237.0, "step": 19420 }, { "epoch": 2.470550820506297, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.067885398864746, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8626610040664673, "num_tokens": 741163593.0, "step": 19421 }, { "epoch": 2.4706780307848875, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.077726364135742, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8635951280593872, "num_tokens": 741202328.0, "step": 19422 }, { "epoch": 2.470805241063478, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.010178565979004, "learning_rate": 1e-06, "loss": 0.3346, "mean_token_accuracy": 0.894047737121582, "num_tokens": 741240095.0, "step": 19423 }, { "epoch": 2.4709324513420685, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.052075386047363, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8760243058204651, "num_tokens": 741278509.0, "step": 19424 }, { "epoch": 2.471059661620659, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.076892852783203, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8698556423187256, "num_tokens": 741321446.0, "step": 19425 }, { "epoch": 2.4711868718992496, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.010534286499023, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8700137138366699, "num_tokens": 741360469.0, "step": 19426 }, { "epoch": 2.47131408217784, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.13101863861084, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8656046390533447, "num_tokens": 741393833.0, "step": 19427 }, { "epoch": 2.4714412924564306, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.029265403747559, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8791632652282715, "num_tokens": 741426956.0, "step": 19428 }, { "epoch": 2.471568502735021, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.114487648010254, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8733375072479248, "num_tokens": 741464820.0, "step": 19429 }, { "epoch": 2.4716957130136117, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.022184371948242, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8818688988685608, "num_tokens": 741499519.0, "step": 19430 }, { "epoch": 2.471822923292202, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.07220458984375, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8814476132392883, "num_tokens": 741533443.0, "step": 19431 }, { "epoch": 2.4719501335707923, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.041122436523438, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8714340925216675, "num_tokens": 741576849.0, "step": 19432 }, { "epoch": 2.4720773438493833, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.024689674377441, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8788480758666992, "num_tokens": 741614653.0, "step": 19433 }, { "epoch": 2.4722045541279734, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.01103401184082, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8693215847015381, "num_tokens": 741649354.0, "step": 19434 }, { "epoch": 2.472331764406564, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.063223838806152, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8672763109207153, "num_tokens": 741692257.0, "step": 19435 }, { "epoch": 2.4724589746851544, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.1347017288208, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8717654943466187, "num_tokens": 741729739.0, "step": 19436 }, { "epoch": 2.472586184963745, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.037604331970215, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8649600744247437, "num_tokens": 741763409.0, "step": 19437 }, { "epoch": 2.4727133952423355, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.061954498291016, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8713483214378357, "num_tokens": 741798746.0, "step": 19438 }, { "epoch": 2.472840605520926, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.083197593688965, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8830248117446899, "num_tokens": 741840488.0, "step": 19439 }, { "epoch": 2.4729678157995165, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.040742874145508, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.8910250663757324, "num_tokens": 741880839.0, "step": 19440 }, { "epoch": 2.473095026078107, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.160253524780273, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8685483336448669, "num_tokens": 741919317.0, "step": 19441 }, { "epoch": 2.4732222363566976, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.077546119689941, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8547967672348022, "num_tokens": 741955782.0, "step": 19442 }, { "epoch": 2.473349446635288, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.141555786132812, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8688024878501892, "num_tokens": 741990410.0, "step": 19443 }, { "epoch": 2.4734766569138786, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.05019474029541, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.868956983089447, "num_tokens": 742031536.0, "step": 19444 }, { "epoch": 2.473603867192469, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.11479663848877, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8637189865112305, "num_tokens": 742067222.0, "step": 19445 }, { "epoch": 2.4737310774710597, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.08643913269043, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8852062225341797, "num_tokens": 742114740.0, "step": 19446 }, { "epoch": 2.47385828774965, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.118678092956543, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.882979154586792, "num_tokens": 742149344.0, "step": 19447 }, { "epoch": 2.4739854980282407, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.165435791015625, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8762212991714478, "num_tokens": 742191089.0, "step": 19448 }, { "epoch": 2.4741127083068313, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.076857566833496, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8743659853935242, "num_tokens": 742224944.0, "step": 19449 }, { "epoch": 2.474239918585422, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.137657165527344, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8637874126434326, "num_tokens": 742263674.0, "step": 19450 }, { "epoch": 2.4743671288640123, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.04822826385498, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8685916066169739, "num_tokens": 742300424.0, "step": 19451 }, { "epoch": 2.474494339142603, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.15269660949707, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8839519619941711, "num_tokens": 742333854.0, "step": 19452 }, { "epoch": 2.4746215494211934, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.079736709594727, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8820914030075073, "num_tokens": 742377126.0, "step": 19453 }, { "epoch": 2.474748759699784, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.127073287963867, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8726249933242798, "num_tokens": 742421284.0, "step": 19454 }, { "epoch": 2.4748759699783744, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.115464210510254, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8578720688819885, "num_tokens": 742459124.0, "step": 19455 }, { "epoch": 2.475003180256965, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.09133529663086, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8647570610046387, "num_tokens": 742493137.0, "step": 19456 }, { "epoch": 2.475130390535555, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.227542877197266, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8770542144775391, "num_tokens": 742529805.0, "step": 19457 }, { "epoch": 2.475257600814146, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.04236125946045, "learning_rate": 1e-06, "loss": 0.3558, "mean_token_accuracy": 0.8867231607437134, "num_tokens": 742560859.0, "step": 19458 }, { "epoch": 2.475384811092736, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.138604164123535, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8863898515701294, "num_tokens": 742601465.0, "step": 19459 }, { "epoch": 2.4755120213713266, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.117827415466309, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8675936460494995, "num_tokens": 742639242.0, "step": 19460 }, { "epoch": 2.475639231649917, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.107078552246094, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8738957643508911, "num_tokens": 742676574.0, "step": 19461 }, { "epoch": 2.4757664419285077, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.115751266479492, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8642117977142334, "num_tokens": 742713326.0, "step": 19462 }, { "epoch": 2.4758936522070982, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.098257064819336, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8510315418243408, "num_tokens": 742756883.0, "step": 19463 }, { "epoch": 2.4760208624856888, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.09714412689209, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8688647747039795, "num_tokens": 742794301.0, "step": 19464 }, { "epoch": 2.4761480727642793, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.072260856628418, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8766171336174011, "num_tokens": 742832966.0, "step": 19465 }, { "epoch": 2.47627528304287, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.075090408325195, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8709345459938049, "num_tokens": 742872926.0, "step": 19466 }, { "epoch": 2.4764024933214603, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.113814353942871, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8812679648399353, "num_tokens": 742910838.0, "step": 19467 }, { "epoch": 2.476529703600051, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.066654205322266, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8827577233314514, "num_tokens": 742950098.0, "step": 19468 }, { "epoch": 2.4766569138786414, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.093117713928223, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.883991539478302, "num_tokens": 742983872.0, "step": 19469 }, { "epoch": 2.476784124157232, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.147785186767578, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8629400134086609, "num_tokens": 743029816.0, "step": 19470 }, { "epoch": 2.4769113344358225, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.075507164001465, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8659151792526245, "num_tokens": 743068565.0, "step": 19471 }, { "epoch": 2.477038544714413, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.126818656921387, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8754802346229553, "num_tokens": 743112137.0, "step": 19472 }, { "epoch": 2.4771657549930035, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.061833381652832, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8880470991134644, "num_tokens": 743145067.0, "step": 19473 }, { "epoch": 2.477292965271594, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.225847244262695, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8782039880752563, "num_tokens": 743177017.0, "step": 19474 }, { "epoch": 2.4774201755501846, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.996293067932129, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.874113917350769, "num_tokens": 743218121.0, "step": 19475 }, { "epoch": 2.477547385828775, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.108867645263672, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8806958198547363, "num_tokens": 743256635.0, "step": 19476 }, { "epoch": 2.4776745961073656, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.073823928833008, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8725813031196594, "num_tokens": 743293801.0, "step": 19477 }, { "epoch": 2.477801806385956, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.1057710647583, "learning_rate": 1e-06, "loss": 0.3536, "mean_token_accuracy": 0.8876800537109375, "num_tokens": 743331582.0, "step": 19478 }, { "epoch": 2.4779290166645467, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.092622756958008, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8777901530265808, "num_tokens": 743369901.0, "step": 19479 }, { "epoch": 2.4780562269431368, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.093467712402344, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8696404099464417, "num_tokens": 743405808.0, "step": 19480 }, { "epoch": 2.4781834372217277, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.060891151428223, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8762775659561157, "num_tokens": 743447236.0, "step": 19481 }, { "epoch": 2.478310647500318, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.002361297607422, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8811130523681641, "num_tokens": 743486494.0, "step": 19482 }, { "epoch": 2.4784378577789083, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.14200210571289, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8707573413848877, "num_tokens": 743523941.0, "step": 19483 }, { "epoch": 2.478565068057499, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.131863594055176, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8793835639953613, "num_tokens": 743554169.0, "step": 19484 }, { "epoch": 2.4786922783360894, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.113835334777832, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.858576238155365, "num_tokens": 743594656.0, "step": 19485 }, { "epoch": 2.47881948861468, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.132773399353027, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8809210062026978, "num_tokens": 743628638.0, "step": 19486 }, { "epoch": 2.4789466988932705, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.078853607177734, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.866233766078949, "num_tokens": 743665382.0, "step": 19487 }, { "epoch": 2.479073909171861, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.10878849029541, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8863813877105713, "num_tokens": 743701712.0, "step": 19488 }, { "epoch": 2.4792011194504515, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.087332725524902, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8742349743843079, "num_tokens": 743740347.0, "step": 19489 }, { "epoch": 2.479328329729042, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.124940872192383, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.87758868932724, "num_tokens": 743775434.0, "step": 19490 }, { "epoch": 2.4794555400076326, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.07752513885498, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8737837672233582, "num_tokens": 743812475.0, "step": 19491 }, { "epoch": 2.479582750286223, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.129097938537598, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8653150796890259, "num_tokens": 743851350.0, "step": 19492 }, { "epoch": 2.4797099605648136, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.075540542602539, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8722372651100159, "num_tokens": 743885086.0, "step": 19493 }, { "epoch": 2.479837170843404, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.047717094421387, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8764941692352295, "num_tokens": 743931747.0, "step": 19494 }, { "epoch": 2.4799643811219947, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.155280113220215, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.88711017370224, "num_tokens": 743971806.0, "step": 19495 }, { "epoch": 2.480091591400585, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.092523574829102, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8845698237419128, "num_tokens": 744005304.0, "step": 19496 }, { "epoch": 2.4802188016791757, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.12652587890625, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8663923740386963, "num_tokens": 744051159.0, "step": 19497 }, { "epoch": 2.4803460119577663, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.053339958190918, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8567747473716736, "num_tokens": 744089856.0, "step": 19498 }, { "epoch": 2.480473222236357, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.099925994873047, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8546910285949707, "num_tokens": 744136355.0, "step": 19499 }, { "epoch": 2.4806004325149473, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.107914924621582, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8636194467544556, "num_tokens": 744177682.0, "step": 19500 }, { "epoch": 2.480727642793538, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.139392852783203, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8753561973571777, "num_tokens": 744214209.0, "step": 19501 }, { "epoch": 2.4808548530721284, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.042190551757812, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8573024272918701, "num_tokens": 744257923.0, "step": 19502 }, { "epoch": 2.480982063350719, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.09152889251709, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.880990743637085, "num_tokens": 744293370.0, "step": 19503 }, { "epoch": 2.4811092736293094, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.155355453491211, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8798642754554749, "num_tokens": 744328572.0, "step": 19504 }, { "epoch": 2.4812364839078995, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.067503929138184, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8583316206932068, "num_tokens": 744364124.0, "step": 19505 }, { "epoch": 2.4813636941864905, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.188329696655273, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8777416944503784, "num_tokens": 744403679.0, "step": 19506 }, { "epoch": 2.4814909044650806, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.103594779968262, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8740319013595581, "num_tokens": 744440922.0, "step": 19507 }, { "epoch": 2.481618114743671, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.099638938903809, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8714789152145386, "num_tokens": 744479025.0, "step": 19508 }, { "epoch": 2.4817453250222616, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.132896423339844, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8762978911399841, "num_tokens": 744514165.0, "step": 19509 }, { "epoch": 2.481872535300852, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.154572486877441, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8695805072784424, "num_tokens": 744547210.0, "step": 19510 }, { "epoch": 2.4819997455794427, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.130990028381348, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8739862442016602, "num_tokens": 744585132.0, "step": 19511 }, { "epoch": 2.4821269558580332, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.088194847106934, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8846531510353088, "num_tokens": 744617921.0, "step": 19512 }, { "epoch": 2.4822541661366238, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.19749927520752, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8683918714523315, "num_tokens": 744653095.0, "step": 19513 }, { "epoch": 2.4823813764152143, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.029279708862305, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8802323341369629, "num_tokens": 744688258.0, "step": 19514 }, { "epoch": 2.482508586693805, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.137503623962402, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8739503026008606, "num_tokens": 744723500.0, "step": 19515 }, { "epoch": 2.4826357969723953, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.060914039611816, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8751435875892639, "num_tokens": 744757133.0, "step": 19516 }, { "epoch": 2.482763007250986, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.110831260681152, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8866397142410278, "num_tokens": 744793814.0, "step": 19517 }, { "epoch": 2.4828902175295764, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.144160270690918, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8782585263252258, "num_tokens": 744828680.0, "step": 19518 }, { "epoch": 2.483017427808167, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.137930870056152, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8687000274658203, "num_tokens": 744861642.0, "step": 19519 }, { "epoch": 2.4831446380867574, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.093986511230469, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8855046033859253, "num_tokens": 744900345.0, "step": 19520 }, { "epoch": 2.483271848365348, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.03524112701416, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.880244791507721, "num_tokens": 744938589.0, "step": 19521 }, { "epoch": 2.4833990586439385, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.196014404296875, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8654448390007019, "num_tokens": 744975385.0, "step": 19522 }, { "epoch": 2.483526268922529, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.088725090026855, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8635125160217285, "num_tokens": 745010860.0, "step": 19523 }, { "epoch": 2.4836534792011196, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.111095428466797, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8665728569030762, "num_tokens": 745047183.0, "step": 19524 }, { "epoch": 2.48378068947971, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.181364059448242, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8713798522949219, "num_tokens": 745081906.0, "step": 19525 }, { "epoch": 2.4839078997583006, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.091506004333496, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8790186643600464, "num_tokens": 745127285.0, "step": 19526 }, { "epoch": 2.484035110036891, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.142223358154297, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.889789879322052, "num_tokens": 745171564.0, "step": 19527 }, { "epoch": 2.4841623203154817, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.103342056274414, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8839949369430542, "num_tokens": 745214252.0, "step": 19528 }, { "epoch": 2.484289530594072, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.072887420654297, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.876594603061676, "num_tokens": 745251983.0, "step": 19529 }, { "epoch": 2.4844167408726623, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.19229507446289, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8754369020462036, "num_tokens": 745287037.0, "step": 19530 }, { "epoch": 2.4845439511512533, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.053754806518555, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8710384368896484, "num_tokens": 745327854.0, "step": 19531 }, { "epoch": 2.4846711614298433, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.113726615905762, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8631598353385925, "num_tokens": 745370227.0, "step": 19532 }, { "epoch": 2.484798371708434, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.08510684967041, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8719261288642883, "num_tokens": 745409416.0, "step": 19533 }, { "epoch": 2.4849255819870244, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.147703170776367, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8570864796638489, "num_tokens": 745449908.0, "step": 19534 }, { "epoch": 2.485052792265615, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.128507614135742, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8636485934257507, "num_tokens": 745486384.0, "step": 19535 }, { "epoch": 2.4851800025442055, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.033333778381348, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8630973696708679, "num_tokens": 745523058.0, "step": 19536 }, { "epoch": 2.485307212822796, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.120796203613281, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8566254377365112, "num_tokens": 745559276.0, "step": 19537 }, { "epoch": 2.4854344231013865, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.094386100769043, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8726708889007568, "num_tokens": 745592726.0, "step": 19538 }, { "epoch": 2.485561633379977, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.059981346130371, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8754235506057739, "num_tokens": 745630117.0, "step": 19539 }, { "epoch": 2.4856888436585676, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.130900382995605, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8604404926300049, "num_tokens": 745664968.0, "step": 19540 }, { "epoch": 2.485816053937158, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.02730655670166, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.875251054763794, "num_tokens": 745701893.0, "step": 19541 }, { "epoch": 2.4859432642157486, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.093399047851562, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8709900379180908, "num_tokens": 745736884.0, "step": 19542 }, { "epoch": 2.486070474494339, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.13417911529541, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8857612609863281, "num_tokens": 745782381.0, "step": 19543 }, { "epoch": 2.4861976847729297, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.088630676269531, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8890309929847717, "num_tokens": 745827630.0, "step": 19544 }, { "epoch": 2.48632489505152, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.126300811767578, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8813996315002441, "num_tokens": 745862223.0, "step": 19545 }, { "epoch": 2.4864521053301107, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.09782886505127, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8888879418373108, "num_tokens": 745903077.0, "step": 19546 }, { "epoch": 2.4865793156087013, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.123906135559082, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8742034435272217, "num_tokens": 745945534.0, "step": 19547 }, { "epoch": 2.486706525887292, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.138898849487305, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8578896522521973, "num_tokens": 745986308.0, "step": 19548 }, { "epoch": 2.4868337361658823, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.111649513244629, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8647683262825012, "num_tokens": 746022206.0, "step": 19549 }, { "epoch": 2.486960946444473, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.105711936950684, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.880352795124054, "num_tokens": 746059854.0, "step": 19550 }, { "epoch": 2.4870881567230634, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.0851411819458, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8813779354095459, "num_tokens": 746103994.0, "step": 19551 }, { "epoch": 2.487215367001654, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.203669548034668, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8640245795249939, "num_tokens": 746138491.0, "step": 19552 }, { "epoch": 2.4873425772802444, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.11474895477295, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8718518018722534, "num_tokens": 746172598.0, "step": 19553 }, { "epoch": 2.487469787558835, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.089967727661133, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8804406523704529, "num_tokens": 746210385.0, "step": 19554 }, { "epoch": 2.487596997837425, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.156899452209473, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8757575154304504, "num_tokens": 746252563.0, "step": 19555 }, { "epoch": 2.487724208116016, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.097907066345215, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8790640830993652, "num_tokens": 746292356.0, "step": 19556 }, { "epoch": 2.487851418394606, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.169621467590332, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8634178638458252, "num_tokens": 746325467.0, "step": 19557 }, { "epoch": 2.4879786286731966, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.127326965332031, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8693522810935974, "num_tokens": 746362285.0, "step": 19558 }, { "epoch": 2.488105838951787, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.168523788452148, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.880908191204071, "num_tokens": 746401540.0, "step": 19559 }, { "epoch": 2.4882330492303777, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.081915855407715, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8749696016311646, "num_tokens": 746439282.0, "step": 19560 }, { "epoch": 2.488360259508968, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.165770530700684, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8847286701202393, "num_tokens": 746480993.0, "step": 19561 }, { "epoch": 2.4884874697875587, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.089710235595703, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8713455200195312, "num_tokens": 746519318.0, "step": 19562 }, { "epoch": 2.4886146800661493, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.222883224487305, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8619053363800049, "num_tokens": 746559761.0, "step": 19563 }, { "epoch": 2.48874189034474, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.103205680847168, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8818085193634033, "num_tokens": 746591622.0, "step": 19564 }, { "epoch": 2.4888691006233303, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.098383903503418, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8732098340988159, "num_tokens": 746627709.0, "step": 19565 }, { "epoch": 2.488996310901921, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.090375900268555, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8712632656097412, "num_tokens": 746663445.0, "step": 19566 }, { "epoch": 2.4891235211805114, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.087919235229492, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8618883490562439, "num_tokens": 746706869.0, "step": 19567 }, { "epoch": 2.489250731459102, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.08240032196045, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8854084610939026, "num_tokens": 746749502.0, "step": 19568 }, { "epoch": 2.4893779417376924, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.090301513671875, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8766509890556335, "num_tokens": 746786510.0, "step": 19569 }, { "epoch": 2.489505152016283, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.190905570983887, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8799035549163818, "num_tokens": 746819430.0, "step": 19570 }, { "epoch": 2.4896323622948735, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.116597175598145, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8768535852432251, "num_tokens": 746856018.0, "step": 19571 }, { "epoch": 2.489759572573464, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.068832397460938, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8684697151184082, "num_tokens": 746897350.0, "step": 19572 }, { "epoch": 2.4898867828520546, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.191304206848145, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8678391575813293, "num_tokens": 746937163.0, "step": 19573 }, { "epoch": 2.490013993130645, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.133042335510254, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.884243905544281, "num_tokens": 746976140.0, "step": 19574 }, { "epoch": 2.4901412034092356, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.086506843566895, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8737202882766724, "num_tokens": 747012576.0, "step": 19575 }, { "epoch": 2.490268413687826, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.129415512084961, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8728514313697815, "num_tokens": 747048179.0, "step": 19576 }, { "epoch": 2.4903956239664167, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.042830467224121, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8552150726318359, "num_tokens": 747085291.0, "step": 19577 }, { "epoch": 2.4905228342450068, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.190695762634277, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8814035654067993, "num_tokens": 747119085.0, "step": 19578 }, { "epoch": 2.4906500445235977, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.047781944274902, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8724120855331421, "num_tokens": 747153463.0, "step": 19579 }, { "epoch": 2.490777254802188, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.170331001281738, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8495705127716064, "num_tokens": 747194781.0, "step": 19580 }, { "epoch": 2.4909044650807783, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.133063316345215, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8749821782112122, "num_tokens": 747235328.0, "step": 19581 }, { "epoch": 2.491031675359369, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.116893768310547, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8710814118385315, "num_tokens": 747278481.0, "step": 19582 }, { "epoch": 2.4911588856379594, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.068859100341797, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8878116607666016, "num_tokens": 747313077.0, "step": 19583 }, { "epoch": 2.49128609591655, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.101704597473145, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.880332887172699, "num_tokens": 747352693.0, "step": 19584 }, { "epoch": 2.4914133061951405, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.021245002746582, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8661189079284668, "num_tokens": 747397717.0, "step": 19585 }, { "epoch": 2.491540516473731, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.116443634033203, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8770848512649536, "num_tokens": 747435457.0, "step": 19586 }, { "epoch": 2.4916677267523215, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.129820823669434, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8784003257751465, "num_tokens": 747479017.0, "step": 19587 }, { "epoch": 2.491794937030912, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.116596221923828, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8734233975410461, "num_tokens": 747520247.0, "step": 19588 }, { "epoch": 2.4919221473095026, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.176130294799805, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8601087331771851, "num_tokens": 747559983.0, "step": 19589 }, { "epoch": 2.492049357588093, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.177268028259277, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8694216012954712, "num_tokens": 747595746.0, "step": 19590 }, { "epoch": 2.4921765678666836, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.0849609375, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8715574741363525, "num_tokens": 747639261.0, "step": 19591 }, { "epoch": 2.492303778145274, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.191550254821777, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8553661108016968, "num_tokens": 747680665.0, "step": 19592 }, { "epoch": 2.4924309884238647, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.129647254943848, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8780860304832458, "num_tokens": 747713320.0, "step": 19593 }, { "epoch": 2.492558198702455, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 14.134941101074219, "learning_rate": 1e-06, "loss": 0.4908, "mean_token_accuracy": 0.8476977348327637, "num_tokens": 747752167.0, "step": 19594 }, { "epoch": 2.4926854089810457, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.087545394897461, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8802381753921509, "num_tokens": 747787274.0, "step": 19595 }, { "epoch": 2.4928126192596363, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.093544960021973, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8677496314048767, "num_tokens": 747826647.0, "step": 19596 }, { "epoch": 2.492939829538227, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.174028396606445, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8698692321777344, "num_tokens": 747866797.0, "step": 19597 }, { "epoch": 2.4930670398168173, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.060966491699219, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.876157283782959, "num_tokens": 747907589.0, "step": 19598 }, { "epoch": 2.493194250095408, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.182561874389648, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8664467334747314, "num_tokens": 747939865.0, "step": 19599 }, { "epoch": 2.4933214603739984, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.05309009552002, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8702813386917114, "num_tokens": 747976238.0, "step": 19600 }, { "epoch": 2.493448670652589, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.111725807189941, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8712696433067322, "num_tokens": 748012076.0, "step": 19601 }, { "epoch": 2.4935758809311794, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.209792137145996, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8678445219993591, "num_tokens": 748042518.0, "step": 19602 }, { "epoch": 2.4937030912097695, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.093233108520508, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8802589178085327, "num_tokens": 748077509.0, "step": 19603 }, { "epoch": 2.4938303014883605, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.128676414489746, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.879051923751831, "num_tokens": 748118164.0, "step": 19604 }, { "epoch": 2.4939575117669506, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.083569526672363, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.867067813873291, "num_tokens": 748153609.0, "step": 19605 }, { "epoch": 2.494084722045541, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.143372535705566, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8712430000305176, "num_tokens": 748189573.0, "step": 19606 }, { "epoch": 2.4942119323241316, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.097209930419922, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8685570955276489, "num_tokens": 748231121.0, "step": 19607 }, { "epoch": 2.494339142602722, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.157063484191895, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8839505314826965, "num_tokens": 748268744.0, "step": 19608 }, { "epoch": 2.4944663528813127, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.149110794067383, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8776348829269409, "num_tokens": 748307441.0, "step": 19609 }, { "epoch": 2.494593563159903, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.131474494934082, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8729630708694458, "num_tokens": 748343506.0, "step": 19610 }, { "epoch": 2.4947207734384937, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.154848098754883, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8743506669998169, "num_tokens": 748382209.0, "step": 19611 }, { "epoch": 2.4948479837170843, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.09636116027832, "learning_rate": 1e-06, "loss": 0.3327, "mean_token_accuracy": 0.8936314582824707, "num_tokens": 748413660.0, "step": 19612 }, { "epoch": 2.494975193995675, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.099295616149902, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8739148378372192, "num_tokens": 748457037.0, "step": 19613 }, { "epoch": 2.4951024042742653, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.212420463562012, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8529461622238159, "num_tokens": 748497227.0, "step": 19614 }, { "epoch": 2.495229614552856, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.112019538879395, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8714892864227295, "num_tokens": 748529901.0, "step": 19615 }, { "epoch": 2.4953568248314464, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.211487770080566, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8603937029838562, "num_tokens": 748568238.0, "step": 19616 }, { "epoch": 2.495484035110037, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.13957405090332, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8633307218551636, "num_tokens": 748608153.0, "step": 19617 }, { "epoch": 2.4956112453886274, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.133304595947266, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8686882257461548, "num_tokens": 748646977.0, "step": 19618 }, { "epoch": 2.495738455667218, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.115883827209473, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8642200231552124, "num_tokens": 748689586.0, "step": 19619 }, { "epoch": 2.4958656659458085, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.169966697692871, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8722977638244629, "num_tokens": 748730915.0, "step": 19620 }, { "epoch": 2.495992876224399, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.143664360046387, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8688619136810303, "num_tokens": 748766996.0, "step": 19621 }, { "epoch": 2.4961200865029896, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.144404411315918, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8663491010665894, "num_tokens": 748801225.0, "step": 19622 }, { "epoch": 2.49624729678158, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.152929306030273, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8679995536804199, "num_tokens": 748841355.0, "step": 19623 }, { "epoch": 2.4963745070601706, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.087847709655762, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8710050582885742, "num_tokens": 748879636.0, "step": 19624 }, { "epoch": 2.496501717338761, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.12675952911377, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.870027482509613, "num_tokens": 748917795.0, "step": 19625 }, { "epoch": 2.4966289276173517, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.110963821411133, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8461825847625732, "num_tokens": 748956512.0, "step": 19626 }, { "epoch": 2.496756137895942, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.116707801818848, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.877065896987915, "num_tokens": 748995577.0, "step": 19627 }, { "epoch": 2.4968833481745323, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.180874824523926, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8690020442008972, "num_tokens": 749027672.0, "step": 19628 }, { "epoch": 2.4970105584531233, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.20741081237793, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8832470178604126, "num_tokens": 749059101.0, "step": 19629 }, { "epoch": 2.4971377687317133, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.06760311126709, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8585634231567383, "num_tokens": 749097034.0, "step": 19630 }, { "epoch": 2.497264979010304, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.133060455322266, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8698761463165283, "num_tokens": 749137536.0, "step": 19631 }, { "epoch": 2.4973921892888944, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.061327934265137, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8731461763381958, "num_tokens": 749171487.0, "step": 19632 }, { "epoch": 2.497519399567485, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.118501663208008, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8649638295173645, "num_tokens": 749213234.0, "step": 19633 }, { "epoch": 2.4976466098460754, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.145225524902344, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8608494997024536, "num_tokens": 749256184.0, "step": 19634 }, { "epoch": 2.497773820124666, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.13673210144043, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8703299164772034, "num_tokens": 749292909.0, "step": 19635 }, { "epoch": 2.4979010304032565, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.172226905822754, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8599945306777954, "num_tokens": 749327947.0, "step": 19636 }, { "epoch": 2.498028240681847, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.132919311523438, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8813872337341309, "num_tokens": 749364107.0, "step": 19637 }, { "epoch": 2.4981554509604376, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.111202239990234, "learning_rate": 1e-06, "loss": 0.3519, "mean_token_accuracy": 0.8877710700035095, "num_tokens": 749398055.0, "step": 19638 }, { "epoch": 2.498282661239028, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.193941116333008, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8716792464256287, "num_tokens": 749434626.0, "step": 19639 }, { "epoch": 2.4984098715176186, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.139958381652832, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8656120300292969, "num_tokens": 749470616.0, "step": 19640 }, { "epoch": 2.498537081796209, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.112598419189453, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.874821126461029, "num_tokens": 749513072.0, "step": 19641 }, { "epoch": 2.4986642920747997, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.095393180847168, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8864985704421997, "num_tokens": 749546988.0, "step": 19642 }, { "epoch": 2.49879150235339, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.020736694335938, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8673973083496094, "num_tokens": 749585851.0, "step": 19643 }, { "epoch": 2.4989187126319807, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.20776653289795, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8708570003509521, "num_tokens": 749619510.0, "step": 19644 }, { "epoch": 2.4990459229105713, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.125523567199707, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8804608583450317, "num_tokens": 749653660.0, "step": 19645 }, { "epoch": 2.499173133189162, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.169642448425293, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8729802370071411, "num_tokens": 749691321.0, "step": 19646 }, { "epoch": 2.4993003434677523, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.062525749206543, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8742567300796509, "num_tokens": 749729305.0, "step": 19647 }, { "epoch": 2.499427553746343, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.131381034851074, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8783469796180725, "num_tokens": 749766134.0, "step": 19648 }, { "epoch": 2.4995547640249334, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.106430053710938, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8697213530540466, "num_tokens": 749807715.0, "step": 19649 }, { "epoch": 2.499681974303524, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.168473243713379, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8737424612045288, "num_tokens": 749838799.0, "step": 19650 }, { "epoch": 2.499809184582114, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.015885353088379, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8857582807540894, "num_tokens": 749874283.0, "step": 19651 }, { "epoch": 2.499936394860705, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.175331115722656, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8700665235519409, "num_tokens": 749915912.0, "step": 19652 }, { "epoch": 2.500063605139295, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.088013648986816, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8789541721343994, "num_tokens": 749959133.0, "step": 19653 }, { "epoch": 2.500190815417886, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.147730827331543, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.869347095489502, "num_tokens": 749997846.0, "step": 19654 }, { "epoch": 2.500318025696476, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.068285942077637, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8764681816101074, "num_tokens": 750040973.0, "step": 19655 }, { "epoch": 2.5004452359750666, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.102982521057129, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8663343787193298, "num_tokens": 750082171.0, "step": 19656 }, { "epoch": 2.500572446253657, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.09047794342041, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.884458065032959, "num_tokens": 750123083.0, "step": 19657 }, { "epoch": 2.5006996565322477, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.121604919433594, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8585106730461121, "num_tokens": 750162235.0, "step": 19658 }, { "epoch": 2.500826866810838, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.124195098876953, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8846556544303894, "num_tokens": 750196501.0, "step": 19659 }, { "epoch": 2.5009540770894287, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.10223388671875, "learning_rate": 1e-06, "loss": 0.4996, "mean_token_accuracy": 0.8442756533622742, "num_tokens": 750241178.0, "step": 19660 }, { "epoch": 2.5010812873680193, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.089550971984863, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8773069977760315, "num_tokens": 750284104.0, "step": 19661 }, { "epoch": 2.50120849764661, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.149190902709961, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8791601061820984, "num_tokens": 750318525.0, "step": 19662 }, { "epoch": 2.5013357079252003, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.146114349365234, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8814314603805542, "num_tokens": 750358892.0, "step": 19663 }, { "epoch": 2.501462918203791, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.146062850952148, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8601987361907959, "num_tokens": 750399325.0, "step": 19664 }, { "epoch": 2.5015901284823814, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.146286010742188, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.865920901298523, "num_tokens": 750437493.0, "step": 19665 }, { "epoch": 2.501717338760972, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.037517547607422, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8801364302635193, "num_tokens": 750472381.0, "step": 19666 }, { "epoch": 2.5018445490395624, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.107617378234863, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8557755947113037, "num_tokens": 750511786.0, "step": 19667 }, { "epoch": 2.501971759318153, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.14219856262207, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8797192573547363, "num_tokens": 750550596.0, "step": 19668 }, { "epoch": 2.5020989695967435, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.086256980895996, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8771334886550903, "num_tokens": 750594790.0, "step": 19669 }, { "epoch": 2.502226179875334, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.158319473266602, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8855319023132324, "num_tokens": 750636738.0, "step": 19670 }, { "epoch": 2.5023533901539246, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.107264518737793, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8852386474609375, "num_tokens": 750672338.0, "step": 19671 }, { "epoch": 2.502480600432515, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.156853675842285, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8668251633644104, "num_tokens": 750709149.0, "step": 19672 }, { "epoch": 2.5026078107111056, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.052478790283203, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8851383924484253, "num_tokens": 750741727.0, "step": 19673 }, { "epoch": 2.5027350209896957, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.17403793334961, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8606491684913635, "num_tokens": 750775903.0, "step": 19674 }, { "epoch": 2.5028622312682867, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.143536567687988, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8691183924674988, "num_tokens": 750810502.0, "step": 19675 }, { "epoch": 2.5029894415468767, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.069517135620117, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8813742399215698, "num_tokens": 750847762.0, "step": 19676 }, { "epoch": 2.5031166518254677, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.201437950134277, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8737090826034546, "num_tokens": 750887201.0, "step": 19677 }, { "epoch": 2.503243862104058, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.076516151428223, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8715920448303223, "num_tokens": 750927323.0, "step": 19678 }, { "epoch": 2.5033710723826488, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.118102073669434, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8645254373550415, "num_tokens": 750972501.0, "step": 19679 }, { "epoch": 2.503498282661239, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.198551177978516, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8832736015319824, "num_tokens": 751010743.0, "step": 19680 }, { "epoch": 2.5036254929398294, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.102874755859375, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8685452342033386, "num_tokens": 751050440.0, "step": 19681 }, { "epoch": 2.50375270321842, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.072432518005371, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8770610094070435, "num_tokens": 751088319.0, "step": 19682 }, { "epoch": 2.5038799134970104, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.151323318481445, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8723881840705872, "num_tokens": 751123766.0, "step": 19683 }, { "epoch": 2.504007123775601, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.134177207946777, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8889856338500977, "num_tokens": 751163375.0, "step": 19684 }, { "epoch": 2.5041343340541915, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.090063095092773, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.868229866027832, "num_tokens": 751204526.0, "step": 19685 }, { "epoch": 2.504261544332782, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.146698951721191, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8780352473258972, "num_tokens": 751245133.0, "step": 19686 }, { "epoch": 2.5043887546113726, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.106779098510742, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8626957535743713, "num_tokens": 751284578.0, "step": 19687 }, { "epoch": 2.504515964889963, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.11309814453125, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8649068474769592, "num_tokens": 751325794.0, "step": 19688 }, { "epoch": 2.5046431751685536, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.101170539855957, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.879589855670929, "num_tokens": 751363777.0, "step": 19689 }, { "epoch": 2.504770385447144, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.09831714630127, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8636594414710999, "num_tokens": 751400499.0, "step": 19690 }, { "epoch": 2.5048975957257347, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.1541166305542, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8709069490432739, "num_tokens": 751434895.0, "step": 19691 }, { "epoch": 2.505024806004325, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.114469528198242, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8821635246276855, "num_tokens": 751473984.0, "step": 19692 }, { "epoch": 2.5051520162829157, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.123612403869629, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8738265037536621, "num_tokens": 751515030.0, "step": 19693 }, { "epoch": 2.5052792265615063, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.134175300598145, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8704814910888672, "num_tokens": 751550023.0, "step": 19694 }, { "epoch": 2.505406436840097, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.063252449035645, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.873630166053772, "num_tokens": 751584793.0, "step": 19695 }, { "epoch": 2.5055336471186873, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.098020553588867, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8550721406936646, "num_tokens": 751619817.0, "step": 19696 }, { "epoch": 2.505660857397278, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.142318725585938, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8791016340255737, "num_tokens": 751657097.0, "step": 19697 }, { "epoch": 2.5057880676758684, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.064854621887207, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.863471269607544, "num_tokens": 751687042.0, "step": 19698 }, { "epoch": 2.5059152779544585, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.07714557647705, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8810747861862183, "num_tokens": 751724281.0, "step": 19699 }, { "epoch": 2.5060424882330494, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.150269508361816, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8538779020309448, "num_tokens": 751765144.0, "step": 19700 }, { "epoch": 2.5061696985116395, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.064536094665527, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8798841238021851, "num_tokens": 751799602.0, "step": 19701 }, { "epoch": 2.5062969087902305, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.102422714233398, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8820827007293701, "num_tokens": 751836972.0, "step": 19702 }, { "epoch": 2.5064241190688206, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.086915016174316, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8746376037597656, "num_tokens": 751869627.0, "step": 19703 }, { "epoch": 2.5065513293474115, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.08569049835205, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8654636144638062, "num_tokens": 751911103.0, "step": 19704 }, { "epoch": 2.5066785396260016, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.085905075073242, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8679794073104858, "num_tokens": 751948956.0, "step": 19705 }, { "epoch": 2.506805749904592, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.07087516784668, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8765490055084229, "num_tokens": 751992185.0, "step": 19706 }, { "epoch": 2.5069329601831827, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.129201889038086, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8695653080940247, "num_tokens": 752032661.0, "step": 19707 }, { "epoch": 2.507060170461773, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.065954208374023, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.877548098564148, "num_tokens": 752073946.0, "step": 19708 }, { "epoch": 2.5071873807403637, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.152517318725586, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8589327335357666, "num_tokens": 752118766.0, "step": 19709 }, { "epoch": 2.5073145910189543, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.109659194946289, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8758077621459961, "num_tokens": 752159669.0, "step": 19710 }, { "epoch": 2.507441801297545, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.078022003173828, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8766971826553345, "num_tokens": 752198139.0, "step": 19711 }, { "epoch": 2.5075690115761353, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.130393028259277, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8750218152999878, "num_tokens": 752239686.0, "step": 19712 }, { "epoch": 2.507696221854726, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.125955581665039, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8556258082389832, "num_tokens": 752275740.0, "step": 19713 }, { "epoch": 2.5078234321333164, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.084491729736328, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.860323965549469, "num_tokens": 752310554.0, "step": 19714 }, { "epoch": 2.507950642411907, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.091532707214355, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8781658411026001, "num_tokens": 752347761.0, "step": 19715 }, { "epoch": 2.5080778526904974, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.03809928894043, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.863002359867096, "num_tokens": 752387128.0, "step": 19716 }, { "epoch": 2.508205062969088, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.106813430786133, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8719727993011475, "num_tokens": 752428722.0, "step": 19717 }, { "epoch": 2.5083322732476785, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.125020027160645, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8712324500083923, "num_tokens": 752468920.0, "step": 19718 }, { "epoch": 2.508459483526269, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.095071792602539, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.863416850566864, "num_tokens": 752505226.0, "step": 19719 }, { "epoch": 2.5085866938048595, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.099844932556152, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8742699027061462, "num_tokens": 752543693.0, "step": 19720 }, { "epoch": 2.50871390408345, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.150348663330078, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8583943843841553, "num_tokens": 752578758.0, "step": 19721 }, { "epoch": 2.5088411143620406, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.067517280578613, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8830440044403076, "num_tokens": 752612115.0, "step": 19722 }, { "epoch": 2.508968324640631, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.1416654586792, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8780847787857056, "num_tokens": 752644601.0, "step": 19723 }, { "epoch": 2.509095534919221, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.143205642700195, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8745906352996826, "num_tokens": 752682333.0, "step": 19724 }, { "epoch": 2.509222745197812, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.10885238647461, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.866497814655304, "num_tokens": 752721802.0, "step": 19725 }, { "epoch": 2.5093499554764023, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.150354385375977, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8878170251846313, "num_tokens": 752755673.0, "step": 19726 }, { "epoch": 2.5094771657549932, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.108999252319336, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.881946861743927, "num_tokens": 752795575.0, "step": 19727 }, { "epoch": 2.5096043760335833, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.126103401184082, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8886687755584717, "num_tokens": 752833172.0, "step": 19728 }, { "epoch": 2.5097315863121743, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.163518905639648, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8582170009613037, "num_tokens": 752872857.0, "step": 19729 }, { "epoch": 2.5098587965907644, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.13794231414795, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8822263479232788, "num_tokens": 752908553.0, "step": 19730 }, { "epoch": 2.509986006869355, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.131742477416992, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8780573606491089, "num_tokens": 752937762.0, "step": 19731 }, { "epoch": 2.5101132171479454, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.129573822021484, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8593524694442749, "num_tokens": 752977184.0, "step": 19732 }, { "epoch": 2.510240427426536, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.099565505981445, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8685281276702881, "num_tokens": 753021071.0, "step": 19733 }, { "epoch": 2.5103676377051265, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.093046188354492, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.8475497961044312, "num_tokens": 753056597.0, "step": 19734 }, { "epoch": 2.510494847983717, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.160054206848145, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8764989972114563, "num_tokens": 753096178.0, "step": 19735 }, { "epoch": 2.5106220582623076, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.148819923400879, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8760512471199036, "num_tokens": 753130388.0, "step": 19736 }, { "epoch": 2.510749268540898, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.100152969360352, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8801854252815247, "num_tokens": 753166585.0, "step": 19737 }, { "epoch": 2.5108764788194886, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.128289222717285, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8819049000740051, "num_tokens": 753203403.0, "step": 19738 }, { "epoch": 2.511003689098079, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.093092918395996, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8787237405776978, "num_tokens": 753241632.0, "step": 19739 }, { "epoch": 2.5111308993766697, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.084253311157227, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.886182963848114, "num_tokens": 753279292.0, "step": 19740 }, { "epoch": 2.51125810965526, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.091787338256836, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8635833859443665, "num_tokens": 753319121.0, "step": 19741 }, { "epoch": 2.5113853199338507, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.088533401489258, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8569893836975098, "num_tokens": 753353074.0, "step": 19742 }, { "epoch": 2.5115125302124413, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.123113632202148, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8910540342330933, "num_tokens": 753388994.0, "step": 19743 }, { "epoch": 2.511639740491032, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.111610412597656, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8793660402297974, "num_tokens": 753430271.0, "step": 19744 }, { "epoch": 2.5117669507696223, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.091731071472168, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8658198118209839, "num_tokens": 753473820.0, "step": 19745 }, { "epoch": 2.511894161048213, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.072304725646973, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8693385720252991, "num_tokens": 753516397.0, "step": 19746 }, { "epoch": 2.5120213713268034, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.134344100952148, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8832488059997559, "num_tokens": 753557589.0, "step": 19747 }, { "epoch": 2.512148581605394, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.16630744934082, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8806236386299133, "num_tokens": 753593764.0, "step": 19748 }, { "epoch": 2.512275791883984, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.172795295715332, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8604733943939209, "num_tokens": 753635967.0, "step": 19749 }, { "epoch": 2.512403002162575, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.145705223083496, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8807322382926941, "num_tokens": 753674595.0, "step": 19750 }, { "epoch": 2.512530212441165, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.145947456359863, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8575214147567749, "num_tokens": 753714118.0, "step": 19751 }, { "epoch": 2.512657422719756, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.23440933227539, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8595049381256104, "num_tokens": 753758139.0, "step": 19752 }, { "epoch": 2.512784632998346, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.074758529663086, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8620532751083374, "num_tokens": 753800373.0, "step": 19753 }, { "epoch": 2.5129118432769366, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.253293991088867, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8675127625465393, "num_tokens": 753832366.0, "step": 19754 }, { "epoch": 2.513039053555527, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.114838600158691, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.891764223575592, "num_tokens": 753866176.0, "step": 19755 }, { "epoch": 2.5131662638341177, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.104175567626953, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8716611266136169, "num_tokens": 753911470.0, "step": 19756 }, { "epoch": 2.513293474112708, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.170701026916504, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8702330589294434, "num_tokens": 753953027.0, "step": 19757 }, { "epoch": 2.5134206843912987, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.11904239654541, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8664790391921997, "num_tokens": 753987992.0, "step": 19758 }, { "epoch": 2.5135478946698893, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.144405364990234, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8777735233306885, "num_tokens": 754026671.0, "step": 19759 }, { "epoch": 2.51367510494848, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.136785507202148, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8733521103858948, "num_tokens": 754063526.0, "step": 19760 }, { "epoch": 2.5138023152270703, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.117136001586914, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8562848567962646, "num_tokens": 754103361.0, "step": 19761 }, { "epoch": 2.513929525505661, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.160917282104492, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8756728172302246, "num_tokens": 754141969.0, "step": 19762 }, { "epoch": 2.5140567357842514, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.122431755065918, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8888564109802246, "num_tokens": 754182691.0, "step": 19763 }, { "epoch": 2.514183946062842, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.150275230407715, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8723816871643066, "num_tokens": 754219956.0, "step": 19764 }, { "epoch": 2.5143111563414324, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.15450668334961, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.869001567363739, "num_tokens": 754259467.0, "step": 19765 }, { "epoch": 2.514438366620023, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.155485153198242, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8760974407196045, "num_tokens": 754300344.0, "step": 19766 }, { "epoch": 2.5145655768986135, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.079245567321777, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8715542554855347, "num_tokens": 754342606.0, "step": 19767 }, { "epoch": 2.514692787177204, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.21096420288086, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8694795370101929, "num_tokens": 754379946.0, "step": 19768 }, { "epoch": 2.5148199974557945, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.118653297424316, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.886721670627594, "num_tokens": 754414353.0, "step": 19769 }, { "epoch": 2.514947207734385, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.149782180786133, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8608297109603882, "num_tokens": 754455506.0, "step": 19770 }, { "epoch": 2.5150744180129756, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.066405296325684, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8561391830444336, "num_tokens": 754498316.0, "step": 19771 }, { "epoch": 2.5152016282915657, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.13620662689209, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8607156276702881, "num_tokens": 754534943.0, "step": 19772 }, { "epoch": 2.5153288385701567, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.019803047180176, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8701474070549011, "num_tokens": 754572775.0, "step": 19773 }, { "epoch": 2.5154560488487467, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.108145713806152, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8829443454742432, "num_tokens": 754610473.0, "step": 19774 }, { "epoch": 2.5155832591273377, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.155564308166504, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8735913038253784, "num_tokens": 754649527.0, "step": 19775 }, { "epoch": 2.515710469405928, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.131635665893555, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8744614124298096, "num_tokens": 754691513.0, "step": 19776 }, { "epoch": 2.5158376796845188, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.122040748596191, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8603337407112122, "num_tokens": 754727932.0, "step": 19777 }, { "epoch": 2.515964889963109, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.196158409118652, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8784956932067871, "num_tokens": 754764571.0, "step": 19778 }, { "epoch": 2.5160921002416994, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.04477310180664, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8865148425102234, "num_tokens": 754805846.0, "step": 19779 }, { "epoch": 2.51621931052029, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.192652702331543, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8794596195220947, "num_tokens": 754844016.0, "step": 19780 }, { "epoch": 2.5163465207988804, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.118454933166504, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8719044327735901, "num_tokens": 754886493.0, "step": 19781 }, { "epoch": 2.516473731077471, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.180046081542969, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8756195306777954, "num_tokens": 754928195.0, "step": 19782 }, { "epoch": 2.5166009413560615, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.157452583312988, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8817583322525024, "num_tokens": 754958399.0, "step": 19783 }, { "epoch": 2.516728151634652, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.186054229736328, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8769004940986633, "num_tokens": 754997926.0, "step": 19784 }, { "epoch": 2.5168553619132426, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.149970054626465, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8788875341415405, "num_tokens": 755040577.0, "step": 19785 }, { "epoch": 2.516982572191833, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.205572128295898, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8705111742019653, "num_tokens": 755070458.0, "step": 19786 }, { "epoch": 2.5171097824704236, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.14875316619873, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8520979881286621, "num_tokens": 755117205.0, "step": 19787 }, { "epoch": 2.517236992749014, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.166631698608398, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8743194937705994, "num_tokens": 755162098.0, "step": 19788 }, { "epoch": 2.5173642030276047, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.14672565460205, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8712671399116516, "num_tokens": 755199655.0, "step": 19789 }, { "epoch": 2.517491413306195, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.246423721313477, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8866267800331116, "num_tokens": 755234165.0, "step": 19790 }, { "epoch": 2.5176186235847857, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.089917182922363, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8748032450675964, "num_tokens": 755271933.0, "step": 19791 }, { "epoch": 2.5177458338633762, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.211723327636719, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8819924592971802, "num_tokens": 755306903.0, "step": 19792 }, { "epoch": 2.5178730441419668, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.157210350036621, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8815738558769226, "num_tokens": 755347613.0, "step": 19793 }, { "epoch": 2.5180002544205573, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.155989646911621, "learning_rate": 1e-06, "loss": 0.3458, "mean_token_accuracy": 0.8920561075210571, "num_tokens": 755380541.0, "step": 19794 }, { "epoch": 2.518127464699148, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.146478652954102, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8712996244430542, "num_tokens": 755422502.0, "step": 19795 }, { "epoch": 2.5182546749777384, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.085739135742188, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8739248514175415, "num_tokens": 755463975.0, "step": 19796 }, { "epoch": 2.5183818852563284, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.201233863830566, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8742731809616089, "num_tokens": 755504873.0, "step": 19797 }, { "epoch": 2.5185090955349194, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.103822708129883, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8487139940261841, "num_tokens": 755545646.0, "step": 19798 }, { "epoch": 2.5186363058135095, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.163990020751953, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8605839610099792, "num_tokens": 755577228.0, "step": 19799 }, { "epoch": 2.5187635160921005, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.148966789245605, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8819462656974792, "num_tokens": 755621437.0, "step": 19800 }, { "epoch": 2.5188907263706906, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.105356216430664, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8709347248077393, "num_tokens": 755658266.0, "step": 19801 }, { "epoch": 2.5190179366492815, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.160829544067383, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8608303070068359, "num_tokens": 755696755.0, "step": 19802 }, { "epoch": 2.5191451469278716, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.141304969787598, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8863562345504761, "num_tokens": 755732434.0, "step": 19803 }, { "epoch": 2.519272357206462, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.121767044067383, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.877642035484314, "num_tokens": 755768488.0, "step": 19804 }, { "epoch": 2.5193995674850527, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.16766357421875, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8796917200088501, "num_tokens": 755799637.0, "step": 19805 }, { "epoch": 2.519526777763643, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.156926155090332, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8579074144363403, "num_tokens": 755834229.0, "step": 19806 }, { "epoch": 2.5196539880422337, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.12867259979248, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8768960237503052, "num_tokens": 755871776.0, "step": 19807 }, { "epoch": 2.5197811983208243, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.075338363647461, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8841347098350525, "num_tokens": 755911125.0, "step": 19808 }, { "epoch": 2.519908408599415, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.14669418334961, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8816642165184021, "num_tokens": 755952204.0, "step": 19809 }, { "epoch": 2.5200356188780053, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.088943481445312, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8621749877929688, "num_tokens": 755996702.0, "step": 19810 }, { "epoch": 2.520162829156596, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.127710342407227, "learning_rate": 1e-06, "loss": 0.352, "mean_token_accuracy": 0.890237033367157, "num_tokens": 756033285.0, "step": 19811 }, { "epoch": 2.5202900394351864, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.122739791870117, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8722151517868042, "num_tokens": 756068851.0, "step": 19812 }, { "epoch": 2.520417249713777, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.150041580200195, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8645204305648804, "num_tokens": 756111676.0, "step": 19813 }, { "epoch": 2.5205444599923674, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.196476936340332, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8578251004219055, "num_tokens": 756146575.0, "step": 19814 }, { "epoch": 2.520671670270958, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.183993339538574, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.86343914270401, "num_tokens": 756190006.0, "step": 19815 }, { "epoch": 2.5207988805495485, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.077590942382812, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8844912648200989, "num_tokens": 756232828.0, "step": 19816 }, { "epoch": 2.520926090828139, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.186651229858398, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8795889616012573, "num_tokens": 756273077.0, "step": 19817 }, { "epoch": 2.5210533011067295, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.121637344360352, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8581146001815796, "num_tokens": 756313421.0, "step": 19818 }, { "epoch": 2.52118051138532, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.14587688446045, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8686019778251648, "num_tokens": 756356461.0, "step": 19819 }, { "epoch": 2.5213077216639106, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.200370788574219, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8698827028274536, "num_tokens": 756388309.0, "step": 19820 }, { "epoch": 2.521434931942501, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.155723571777344, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8841246366500854, "num_tokens": 756426655.0, "step": 19821 }, { "epoch": 2.521562142221091, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.181061744689941, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.870205283164978, "num_tokens": 756464961.0, "step": 19822 }, { "epoch": 2.521689352499682, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.106621742248535, "learning_rate": 1e-06, "loss": 0.3456, "mean_token_accuracy": 0.8881115317344666, "num_tokens": 756497338.0, "step": 19823 }, { "epoch": 2.5218165627782723, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.186993598937988, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8795269727706909, "num_tokens": 756535324.0, "step": 19824 }, { "epoch": 2.5219437730568632, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.079987525939941, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8817666172981262, "num_tokens": 756574138.0, "step": 19825 }, { "epoch": 2.5220709833354533, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.120985984802246, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8830879926681519, "num_tokens": 756608967.0, "step": 19826 }, { "epoch": 2.522198193614044, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.13180923461914, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8525350689888, "num_tokens": 756653882.0, "step": 19827 }, { "epoch": 2.5223254038926344, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.188826560974121, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8684706687927246, "num_tokens": 756695638.0, "step": 19828 }, { "epoch": 2.522452614171225, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.074592590332031, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.866906464099884, "num_tokens": 756734964.0, "step": 19829 }, { "epoch": 2.5225798244498154, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.143664360046387, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8781770467758179, "num_tokens": 756768454.0, "step": 19830 }, { "epoch": 2.522707034728406, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.17204761505127, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8630006909370422, "num_tokens": 756801696.0, "step": 19831 }, { "epoch": 2.5228342450069965, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.151054382324219, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.874178409576416, "num_tokens": 756839874.0, "step": 19832 }, { "epoch": 2.522961455285587, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.152745246887207, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8626917600631714, "num_tokens": 756884179.0, "step": 19833 }, { "epoch": 2.5230886655641775, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.109251976013184, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8752077221870422, "num_tokens": 756922725.0, "step": 19834 }, { "epoch": 2.523215875842768, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.183547019958496, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8688220977783203, "num_tokens": 756956080.0, "step": 19835 }, { "epoch": 2.5233430861213586, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.149014472961426, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8635635375976562, "num_tokens": 756995137.0, "step": 19836 }, { "epoch": 2.523470296399949, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.207719802856445, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8792955875396729, "num_tokens": 757035300.0, "step": 19837 }, { "epoch": 2.5235975066785397, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.063369750976562, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8501110076904297, "num_tokens": 757068789.0, "step": 19838 }, { "epoch": 2.52372471695713, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.16751480102539, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8681837916374207, "num_tokens": 757102913.0, "step": 19839 }, { "epoch": 2.5238519272357207, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.083971977233887, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8742965459823608, "num_tokens": 757138932.0, "step": 19840 }, { "epoch": 2.5239791375143112, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.14297103881836, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8756536841392517, "num_tokens": 757179181.0, "step": 19841 }, { "epoch": 2.5241063477929018, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.123225212097168, "learning_rate": 1e-06, "loss": 0.4895, "mean_token_accuracy": 0.8429653644561768, "num_tokens": 757224209.0, "step": 19842 }, { "epoch": 2.5242335580714923, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.116601943969727, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8859938979148865, "num_tokens": 757262398.0, "step": 19843 }, { "epoch": 2.524360768350083, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.198110580444336, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8628532886505127, "num_tokens": 757301744.0, "step": 19844 }, { "epoch": 2.5244879786286734, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.205880165100098, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.877498984336853, "num_tokens": 757338273.0, "step": 19845 }, { "epoch": 2.524615188907264, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.132457733154297, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8652549982070923, "num_tokens": 757383169.0, "step": 19846 }, { "epoch": 2.524742399185854, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.172450065612793, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8618802428245544, "num_tokens": 757420998.0, "step": 19847 }, { "epoch": 2.524869609464445, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.115242004394531, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.877623438835144, "num_tokens": 757463281.0, "step": 19848 }, { "epoch": 2.524996819743035, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.192423820495605, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8674001693725586, "num_tokens": 757495022.0, "step": 19849 }, { "epoch": 2.525124030021626, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.070645332336426, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8783122301101685, "num_tokens": 757536801.0, "step": 19850 }, { "epoch": 2.525251240300216, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.159668922424316, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8872526288032532, "num_tokens": 757573419.0, "step": 19851 }, { "epoch": 2.5253784505788066, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.21603775024414, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8675191402435303, "num_tokens": 757610773.0, "step": 19852 }, { "epoch": 2.525505660857397, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.154032707214355, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8829389810562134, "num_tokens": 757648067.0, "step": 19853 }, { "epoch": 2.5256328711359877, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.082059860229492, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8789440393447876, "num_tokens": 757681598.0, "step": 19854 }, { "epoch": 2.525760081414578, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.19141960144043, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8548131585121155, "num_tokens": 757724752.0, "step": 19855 }, { "epoch": 2.5258872916931687, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.073159217834473, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8842788338661194, "num_tokens": 757763801.0, "step": 19856 }, { "epoch": 2.5260145019717593, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.161107063293457, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8712941408157349, "num_tokens": 757799235.0, "step": 19857 }, { "epoch": 2.52614171225035, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.1423978805542, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8768368363380432, "num_tokens": 757839695.0, "step": 19858 }, { "epoch": 2.5262689225289403, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.127286911010742, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8598523139953613, "num_tokens": 757877152.0, "step": 19859 }, { "epoch": 2.526396132807531, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.131684303283691, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8719786405563354, "num_tokens": 757916173.0, "step": 19860 }, { "epoch": 2.5265233430861214, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.09036922454834, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8774154782295227, "num_tokens": 757957271.0, "step": 19861 }, { "epoch": 2.526650553364712, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.1400146484375, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8671509027481079, "num_tokens": 757995553.0, "step": 19862 }, { "epoch": 2.5267777636433024, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.176972389221191, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8619821667671204, "num_tokens": 758035095.0, "step": 19863 }, { "epoch": 2.526904973921893, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.099535942077637, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8707437515258789, "num_tokens": 758071185.0, "step": 19864 }, { "epoch": 2.5270321842004835, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.155757904052734, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8737807273864746, "num_tokens": 758110251.0, "step": 19865 }, { "epoch": 2.527159394479074, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.149657249450684, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8587730526924133, "num_tokens": 758144385.0, "step": 19866 }, { "epoch": 2.5272866047576645, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.133918762207031, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8843160271644592, "num_tokens": 758185935.0, "step": 19867 }, { "epoch": 2.527413815036255, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.084312438964844, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8764415979385376, "num_tokens": 758226688.0, "step": 19868 }, { "epoch": 2.5275410253148456, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.121630668640137, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8786885142326355, "num_tokens": 758263866.0, "step": 19869 }, { "epoch": 2.5276682355934357, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.12651538848877, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8774400353431702, "num_tokens": 758303607.0, "step": 19870 }, { "epoch": 2.5277954458720266, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.105626106262207, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8769914507865906, "num_tokens": 758340120.0, "step": 19871 }, { "epoch": 2.5279226561506167, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.141341209411621, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8732115030288696, "num_tokens": 758373792.0, "step": 19872 }, { "epoch": 2.5280498664292077, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.18438720703125, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8615463972091675, "num_tokens": 758418157.0, "step": 19873 }, { "epoch": 2.528177076707798, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.11040210723877, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8652160167694092, "num_tokens": 758454042.0, "step": 19874 }, { "epoch": 2.5283042869863888, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.167447090148926, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8742328882217407, "num_tokens": 758495305.0, "step": 19875 }, { "epoch": 2.528431497264979, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.114535331726074, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8803947567939758, "num_tokens": 758532646.0, "step": 19876 }, { "epoch": 2.5285587075435694, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.176660537719727, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.868038535118103, "num_tokens": 758568329.0, "step": 19877 }, { "epoch": 2.52868591782216, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.116750717163086, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8757047653198242, "num_tokens": 758608246.0, "step": 19878 }, { "epoch": 2.5288131281007504, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.147967338562012, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8821306824684143, "num_tokens": 758645356.0, "step": 19879 }, { "epoch": 2.528940338379341, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.139984130859375, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8875445127487183, "num_tokens": 758685657.0, "step": 19880 }, { "epoch": 2.5290675486579315, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.179819107055664, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8765998482704163, "num_tokens": 758718843.0, "step": 19881 }, { "epoch": 2.529194758936522, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.149399757385254, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8765861988067627, "num_tokens": 758757946.0, "step": 19882 }, { "epoch": 2.5293219692151125, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.18049144744873, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8543100357055664, "num_tokens": 758792832.0, "step": 19883 }, { "epoch": 2.529449179493703, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.082205772399902, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8658185005187988, "num_tokens": 758826062.0, "step": 19884 }, { "epoch": 2.5295763897722936, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.164668083190918, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8815616369247437, "num_tokens": 758860940.0, "step": 19885 }, { "epoch": 2.529703600050884, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.115760803222656, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8724352121353149, "num_tokens": 758900740.0, "step": 19886 }, { "epoch": 2.5298308103294747, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.105271339416504, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8694635033607483, "num_tokens": 758942786.0, "step": 19887 }, { "epoch": 2.529958020608065, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.099818229675293, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8815869688987732, "num_tokens": 758983376.0, "step": 19888 }, { "epoch": 2.5300852308866557, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.162941932678223, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8623396158218384, "num_tokens": 759024007.0, "step": 19889 }, { "epoch": 2.5302124411652462, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.059854507446289, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8799660205841064, "num_tokens": 759061611.0, "step": 19890 }, { "epoch": 2.5303396514438368, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.132030487060547, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8568050265312195, "num_tokens": 759101942.0, "step": 19891 }, { "epoch": 2.5304668617224273, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.111323356628418, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.889640212059021, "num_tokens": 759140281.0, "step": 19892 }, { "epoch": 2.530594072001018, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.19920825958252, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8726155757904053, "num_tokens": 759174319.0, "step": 19893 }, { "epoch": 2.5307212822796084, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.243973731994629, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8798695802688599, "num_tokens": 759208981.0, "step": 19894 }, { "epoch": 2.5308484925581984, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.22203540802002, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8747601509094238, "num_tokens": 759240559.0, "step": 19895 }, { "epoch": 2.5309757028367894, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.141953468322754, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.885021984577179, "num_tokens": 759280550.0, "step": 19896 }, { "epoch": 2.5311029131153795, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.130945205688477, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8667978644371033, "num_tokens": 759321298.0, "step": 19897 }, { "epoch": 2.5312301233939705, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.142436981201172, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.885044515132904, "num_tokens": 759354627.0, "step": 19898 }, { "epoch": 2.5313573336725606, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.192683219909668, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8683591485023499, "num_tokens": 759388129.0, "step": 19899 }, { "epoch": 2.5314845439511515, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.169074058532715, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8817268013954163, "num_tokens": 759423667.0, "step": 19900 }, { "epoch": 2.5316117542297416, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.116683959960938, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8639785051345825, "num_tokens": 759466586.0, "step": 19901 }, { "epoch": 2.531738964508332, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.072761535644531, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8674564957618713, "num_tokens": 759504214.0, "step": 19902 }, { "epoch": 2.5318661747869227, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.156195640563965, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8597660064697266, "num_tokens": 759543865.0, "step": 19903 }, { "epoch": 2.531993385065513, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.078596115112305, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8690338730812073, "num_tokens": 759581917.0, "step": 19904 }, { "epoch": 2.5321205953441037, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.185907363891602, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8759998083114624, "num_tokens": 759617641.0, "step": 19905 }, { "epoch": 2.5322478056226942, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.174378395080566, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8824824094772339, "num_tokens": 759649097.0, "step": 19906 }, { "epoch": 2.5323750159012848, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.140149116516113, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8684027194976807, "num_tokens": 759688802.0, "step": 19907 }, { "epoch": 2.5325022261798753, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.074997901916504, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8827000856399536, "num_tokens": 759724243.0, "step": 19908 }, { "epoch": 2.532629436458466, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.11624813079834, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8664266467094421, "num_tokens": 759764646.0, "step": 19909 }, { "epoch": 2.5327566467370564, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.155710220336914, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8754494190216064, "num_tokens": 759806939.0, "step": 19910 }, { "epoch": 2.532883857015647, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.19106388092041, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8753454685211182, "num_tokens": 759848138.0, "step": 19911 }, { "epoch": 2.5330110672942374, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.128145217895508, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8701068162918091, "num_tokens": 759881877.0, "step": 19912 }, { "epoch": 2.533138277572828, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.139934539794922, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8832781314849854, "num_tokens": 759922625.0, "step": 19913 }, { "epoch": 2.5332654878514185, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.132698059082031, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8846737146377563, "num_tokens": 759952429.0, "step": 19914 }, { "epoch": 2.533392698130009, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.198237419128418, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8754609823226929, "num_tokens": 759987678.0, "step": 19915 }, { "epoch": 2.5335199084085995, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.140774726867676, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8598008751869202, "num_tokens": 760023943.0, "step": 19916 }, { "epoch": 2.53364711868719, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.153543472290039, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8761985301971436, "num_tokens": 760068229.0, "step": 19917 }, { "epoch": 2.5337743289657806, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.171239852905273, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8603366017341614, "num_tokens": 760109440.0, "step": 19918 }, { "epoch": 2.533901539244371, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.129119873046875, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8746432065963745, "num_tokens": 760146616.0, "step": 19919 }, { "epoch": 2.534028749522961, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.159518241882324, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8558265566825867, "num_tokens": 760188708.0, "step": 19920 }, { "epoch": 2.534155959801552, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.107368469238281, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8893329501152039, "num_tokens": 760231355.0, "step": 19921 }, { "epoch": 2.5342831700801423, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.169092178344727, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.861717939376831, "num_tokens": 760268316.0, "step": 19922 }, { "epoch": 2.5344103803587332, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.145767211914062, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.864497184753418, "num_tokens": 760305420.0, "step": 19923 }, { "epoch": 2.5345375906373233, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.115556716918945, "learning_rate": 1e-06, "loss": 0.3423, "mean_token_accuracy": 0.8914887309074402, "num_tokens": 760342195.0, "step": 19924 }, { "epoch": 2.534664800915914, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.19849967956543, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8869572877883911, "num_tokens": 760375394.0, "step": 19925 }, { "epoch": 2.5347920111945044, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.13901138305664, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8617282509803772, "num_tokens": 760411629.0, "step": 19926 }, { "epoch": 2.534919221473095, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.123907089233398, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8757642507553101, "num_tokens": 760456212.0, "step": 19927 }, { "epoch": 2.5350464317516854, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.144505500793457, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8638654351234436, "num_tokens": 760496385.0, "step": 19928 }, { "epoch": 2.535173642030276, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.187865257263184, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8724710941314697, "num_tokens": 760530426.0, "step": 19929 }, { "epoch": 2.5353008523088665, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.135843276977539, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8844245076179504, "num_tokens": 760568174.0, "step": 19930 }, { "epoch": 2.535428062587457, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.097369194030762, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8914405107498169, "num_tokens": 760608097.0, "step": 19931 }, { "epoch": 2.5355552728660475, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.176039695739746, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8731549978256226, "num_tokens": 760643018.0, "step": 19932 }, { "epoch": 2.535682483144638, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.082995414733887, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8809283971786499, "num_tokens": 760676553.0, "step": 19933 }, { "epoch": 2.5358096934232286, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.131918907165527, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8877390027046204, "num_tokens": 760708850.0, "step": 19934 }, { "epoch": 2.535936903701819, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.096668243408203, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8698166608810425, "num_tokens": 760746088.0, "step": 19935 }, { "epoch": 2.5360641139804097, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.127184867858887, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8786330819129944, "num_tokens": 760785716.0, "step": 19936 }, { "epoch": 2.536191324259, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.160499572753906, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8823729753494263, "num_tokens": 760821301.0, "step": 19937 }, { "epoch": 2.5363185345375907, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.113919258117676, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8626094460487366, "num_tokens": 760861200.0, "step": 19938 }, { "epoch": 2.5364457448161812, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.113637924194336, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8723334074020386, "num_tokens": 760896398.0, "step": 19939 }, { "epoch": 2.5365729550947718, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.178391456604004, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8541414737701416, "num_tokens": 760937300.0, "step": 19940 }, { "epoch": 2.5367001653733623, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.202169418334961, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8684117794036865, "num_tokens": 760973726.0, "step": 19941 }, { "epoch": 2.536827375651953, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.132624626159668, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8717668056488037, "num_tokens": 761009048.0, "step": 19942 }, { "epoch": 2.5369545859305433, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.128838539123535, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.871386706829071, "num_tokens": 761050836.0, "step": 19943 }, { "epoch": 2.537081796209134, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.124772071838379, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8667033910751343, "num_tokens": 761087517.0, "step": 19944 }, { "epoch": 2.537209006487724, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.156793594360352, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8869400024414062, "num_tokens": 761121401.0, "step": 19945 }, { "epoch": 2.537336216766315, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.221864700317383, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8628504276275635, "num_tokens": 761154973.0, "step": 19946 }, { "epoch": 2.537463427044905, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.071612358093262, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8814725875854492, "num_tokens": 761194211.0, "step": 19947 }, { "epoch": 2.537590637323496, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.211858749389648, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8665865063667297, "num_tokens": 761233486.0, "step": 19948 }, { "epoch": 2.537717847602086, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.14699649810791, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8743941783905029, "num_tokens": 761275846.0, "step": 19949 }, { "epoch": 2.5378450578806766, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.110559463500977, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8634184002876282, "num_tokens": 761311210.0, "step": 19950 }, { "epoch": 2.537972268159267, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.191051483154297, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8723208904266357, "num_tokens": 761349961.0, "step": 19951 }, { "epoch": 2.5380994784378577, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.189596176147461, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8800663352012634, "num_tokens": 761386993.0, "step": 19952 }, { "epoch": 2.538226688716448, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.181159973144531, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8630186319351196, "num_tokens": 761426910.0, "step": 19953 }, { "epoch": 2.5383538989950387, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.158611297607422, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8751947283744812, "num_tokens": 761466633.0, "step": 19954 }, { "epoch": 2.5384811092736292, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.202958106994629, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.860434889793396, "num_tokens": 761497670.0, "step": 19955 }, { "epoch": 2.5386083195522198, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.195547103881836, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8702370524406433, "num_tokens": 761534626.0, "step": 19956 }, { "epoch": 2.5387355298308103, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.13807487487793, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.880713701248169, "num_tokens": 761581125.0, "step": 19957 }, { "epoch": 2.538862740109401, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.22083568572998, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8740335702896118, "num_tokens": 761616533.0, "step": 19958 }, { "epoch": 2.5389899503879914, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.132330894470215, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8632249236106873, "num_tokens": 761650615.0, "step": 19959 }, { "epoch": 2.539117160666582, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.221545219421387, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8836233615875244, "num_tokens": 761687020.0, "step": 19960 }, { "epoch": 2.5392443709451724, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.21415901184082, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8891919851303101, "num_tokens": 761721653.0, "step": 19961 }, { "epoch": 2.539371581223763, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.21491813659668, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8805792331695557, "num_tokens": 761758377.0, "step": 19962 }, { "epoch": 2.5394987915023535, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.123900413513184, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8732120394706726, "num_tokens": 761795263.0, "step": 19963 }, { "epoch": 2.539626001780944, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.199925422668457, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8831608295440674, "num_tokens": 761834758.0, "step": 19964 }, { "epoch": 2.5397532120595345, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.159375190734863, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8664094805717468, "num_tokens": 761880482.0, "step": 19965 }, { "epoch": 2.539880422338125, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.081729888916016, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8723772764205933, "num_tokens": 761922494.0, "step": 19966 }, { "epoch": 2.5400076326167156, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.262259483337402, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.8909826874732971, "num_tokens": 761959462.0, "step": 19967 }, { "epoch": 2.5401348428953057, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.247556686401367, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8643320202827454, "num_tokens": 761990316.0, "step": 19968 }, { "epoch": 2.5402620531738966, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.182369232177734, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8679683804512024, "num_tokens": 762023671.0, "step": 19969 }, { "epoch": 2.5403892634524867, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.256607055664062, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.878626823425293, "num_tokens": 762056719.0, "step": 19970 }, { "epoch": 2.5405164737310777, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.141801834106445, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8742545247077942, "num_tokens": 762098802.0, "step": 19971 }, { "epoch": 2.540643684009668, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.252114295959473, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.875969409942627, "num_tokens": 762134470.0, "step": 19972 }, { "epoch": 2.5407708942882588, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.119256019592285, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8759032487869263, "num_tokens": 762177163.0, "step": 19973 }, { "epoch": 2.540898104566849, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.228023529052734, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8811436891555786, "num_tokens": 762217323.0, "step": 19974 }, { "epoch": 2.5410253148454394, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.148102760314941, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8699941635131836, "num_tokens": 762252144.0, "step": 19975 }, { "epoch": 2.54115252512403, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.28659439086914, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8664675354957581, "num_tokens": 762282108.0, "step": 19976 }, { "epoch": 2.5412797354026204, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.147149085998535, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8765244483947754, "num_tokens": 762318154.0, "step": 19977 }, { "epoch": 2.541406945681211, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.191754341125488, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8704612255096436, "num_tokens": 762354704.0, "step": 19978 }, { "epoch": 2.5415341559598015, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.11618423461914, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8757625222206116, "num_tokens": 762389943.0, "step": 19979 }, { "epoch": 2.541661366238392, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.165644645690918, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8757482767105103, "num_tokens": 762421560.0, "step": 19980 }, { "epoch": 2.5417885765169825, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.133551597595215, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8792115449905396, "num_tokens": 762458509.0, "step": 19981 }, { "epoch": 2.541915786795573, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.073172569274902, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8725478649139404, "num_tokens": 762502382.0, "step": 19982 }, { "epoch": 2.5420429970741636, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.19477653503418, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8694866299629211, "num_tokens": 762536611.0, "step": 19983 }, { "epoch": 2.542170207352754, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.123686790466309, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8737428188323975, "num_tokens": 762577801.0, "step": 19984 }, { "epoch": 2.5422974176313446, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.135887145996094, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8759208917617798, "num_tokens": 762617413.0, "step": 19985 }, { "epoch": 2.542424627909935, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.13973617553711, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.882220983505249, "num_tokens": 762654811.0, "step": 19986 }, { "epoch": 2.5425518381885257, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.106077194213867, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8793467283248901, "num_tokens": 762692902.0, "step": 19987 }, { "epoch": 2.5426790484671162, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.18756103515625, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8740248680114746, "num_tokens": 762731984.0, "step": 19988 }, { "epoch": 2.5428062587457068, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.12137508392334, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8823697566986084, "num_tokens": 762767104.0, "step": 19989 }, { "epoch": 2.5429334690242973, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.116744041442871, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8618413209915161, "num_tokens": 762804982.0, "step": 19990 }, { "epoch": 2.543060679302888, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.13920783996582, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8630056977272034, "num_tokens": 762842956.0, "step": 19991 }, { "epoch": 2.5431878895814783, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.215633392333984, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8792933225631714, "num_tokens": 762875445.0, "step": 19992 }, { "epoch": 2.5433150998600684, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.183418273925781, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8875163793563843, "num_tokens": 762906061.0, "step": 19993 }, { "epoch": 2.5434423101386594, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.157958030700684, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8814102411270142, "num_tokens": 762945550.0, "step": 19994 }, { "epoch": 2.5435695204172495, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.176786422729492, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8735084533691406, "num_tokens": 762982586.0, "step": 19995 }, { "epoch": 2.5436967306958405, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.205756187438965, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8742684721946716, "num_tokens": 763013527.0, "step": 19996 }, { "epoch": 2.5438239409744305, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.264131546020508, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8686146140098572, "num_tokens": 763054363.0, "step": 19997 }, { "epoch": 2.5439511512530215, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.137621879577637, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8690629601478577, "num_tokens": 763096834.0, "step": 19998 }, { "epoch": 2.5440783615316116, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.197979927062988, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8723544478416443, "num_tokens": 763139611.0, "step": 19999 }, { "epoch": 2.544205571810202, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.299565315246582, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8748286962509155, "num_tokens": 763181085.0, "step": 20000 }, { "epoch": 2.5443327820887927, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.268226623535156, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8709398508071899, "num_tokens": 763219811.0, "step": 20001 }, { "epoch": 2.544459992367383, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.2158842086792, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.862569272518158, "num_tokens": 763257745.0, "step": 20002 }, { "epoch": 2.5445872026459737, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.214845657348633, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8580920696258545, "num_tokens": 763299475.0, "step": 20003 }, { "epoch": 2.5447144129245642, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.225462913513184, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8524437546730042, "num_tokens": 763333960.0, "step": 20004 }, { "epoch": 2.5448416232031548, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.188752174377441, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8645982146263123, "num_tokens": 763378646.0, "step": 20005 }, { "epoch": 2.5449688334817453, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.150703430175781, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8778342008590698, "num_tokens": 763412441.0, "step": 20006 }, { "epoch": 2.545096043760336, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.167201042175293, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.877652645111084, "num_tokens": 763446713.0, "step": 20007 }, { "epoch": 2.5452232540389264, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.146902084350586, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.882763147354126, "num_tokens": 763478686.0, "step": 20008 }, { "epoch": 2.545350464317517, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.153007507324219, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8876707553863525, "num_tokens": 763517258.0, "step": 20009 }, { "epoch": 2.5454776745961074, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.199164390563965, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8684737682342529, "num_tokens": 763553229.0, "step": 20010 }, { "epoch": 2.545604884874698, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.169045448303223, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8804911375045776, "num_tokens": 763594274.0, "step": 20011 }, { "epoch": 2.5457320951532885, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.249088287353516, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8655823469161987, "num_tokens": 763632162.0, "step": 20012 }, { "epoch": 2.545859305431879, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.200254440307617, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8739370107650757, "num_tokens": 763666797.0, "step": 20013 }, { "epoch": 2.5459865157104695, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.250082015991211, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.871030330657959, "num_tokens": 763705475.0, "step": 20014 }, { "epoch": 2.54611372598906, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.22771167755127, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8893183469772339, "num_tokens": 763739434.0, "step": 20015 }, { "epoch": 2.5462409362676506, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.147029876708984, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.861604630947113, "num_tokens": 763773262.0, "step": 20016 }, { "epoch": 2.546368146546241, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.210524559020996, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8624197244644165, "num_tokens": 763812573.0, "step": 20017 }, { "epoch": 2.546495356824831, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.147134780883789, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.876342236995697, "num_tokens": 763848555.0, "step": 20018 }, { "epoch": 2.546622567103422, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.219457626342773, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8716058135032654, "num_tokens": 763884608.0, "step": 20019 }, { "epoch": 2.5467497773820122, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.128632545471191, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.869308352470398, "num_tokens": 763922038.0, "step": 20020 }, { "epoch": 2.546876987660603, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.18935489654541, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.870476484298706, "num_tokens": 763958089.0, "step": 20021 }, { "epoch": 2.5470041979391933, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.180397033691406, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8739819526672363, "num_tokens": 764000802.0, "step": 20022 }, { "epoch": 2.547131408217784, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.201637268066406, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8774338960647583, "num_tokens": 764036628.0, "step": 20023 }, { "epoch": 2.5472586184963744, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.13728141784668, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8761603832244873, "num_tokens": 764073643.0, "step": 20024 }, { "epoch": 2.547385828774965, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.183307647705078, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8795351386070251, "num_tokens": 764110473.0, "step": 20025 }, { "epoch": 2.5475130390535554, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.222952842712402, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8725853562355042, "num_tokens": 764142941.0, "step": 20026 }, { "epoch": 2.547640249332146, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.179577827453613, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.875664472579956, "num_tokens": 764178752.0, "step": 20027 }, { "epoch": 2.5477674596107365, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.277416229248047, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8707668781280518, "num_tokens": 764217440.0, "step": 20028 }, { "epoch": 2.547894669889327, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.138989448547363, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8643573522567749, "num_tokens": 764257524.0, "step": 20029 }, { "epoch": 2.5480218801679175, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.185279846191406, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8696809411048889, "num_tokens": 764293645.0, "step": 20030 }, { "epoch": 2.548149090446508, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.140618324279785, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.880683958530426, "num_tokens": 764331660.0, "step": 20031 }, { "epoch": 2.5482763007250986, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.21481990814209, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.8902289271354675, "num_tokens": 764376577.0, "step": 20032 }, { "epoch": 2.548403511003689, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.150216102600098, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8711714744567871, "num_tokens": 764415492.0, "step": 20033 }, { "epoch": 2.5485307212822796, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.225395202636719, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8642263412475586, "num_tokens": 764453948.0, "step": 20034 }, { "epoch": 2.54865793156087, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.122618675231934, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8777642846107483, "num_tokens": 764493906.0, "step": 20035 }, { "epoch": 2.5487851418394607, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.207441329956055, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8670556545257568, "num_tokens": 764532379.0, "step": 20036 }, { "epoch": 2.5489123521180512, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.235284805297852, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8552684783935547, "num_tokens": 764567726.0, "step": 20037 }, { "epoch": 2.5490395623966418, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.141633033752441, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8740364909172058, "num_tokens": 764606229.0, "step": 20038 }, { "epoch": 2.5491667726752323, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.196760177612305, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8834352493286133, "num_tokens": 764641346.0, "step": 20039 }, { "epoch": 2.549293982953823, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.058846473693848, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8742127418518066, "num_tokens": 764687286.0, "step": 20040 }, { "epoch": 2.5494211932324133, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.103414535522461, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8768154382705688, "num_tokens": 764729235.0, "step": 20041 }, { "epoch": 2.549548403511004, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.169898986816406, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8648828268051147, "num_tokens": 764771834.0, "step": 20042 }, { "epoch": 2.549675613789594, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.118440628051758, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8735737800598145, "num_tokens": 764808307.0, "step": 20043 }, { "epoch": 2.549802824068185, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.1560697555542, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.866333544254303, "num_tokens": 764848373.0, "step": 20044 }, { "epoch": 2.549930034346775, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.08242130279541, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8699886798858643, "num_tokens": 764886468.0, "step": 20045 }, { "epoch": 2.550057244625366, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.175232887268066, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.875301718711853, "num_tokens": 764928690.0, "step": 20046 }, { "epoch": 2.550184454903956, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.107852935791016, "learning_rate": 1e-06, "loss": 0.5078, "mean_token_accuracy": 0.8398807048797607, "num_tokens": 764969148.0, "step": 20047 }, { "epoch": 2.5503116651825466, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.112926483154297, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8669828176498413, "num_tokens": 765010754.0, "step": 20048 }, { "epoch": 2.550438875461137, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.087479591369629, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8766462802886963, "num_tokens": 765043562.0, "step": 20049 }, { "epoch": 2.5505660857397277, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.117345809936523, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8541451096534729, "num_tokens": 765085638.0, "step": 20050 }, { "epoch": 2.550693296018318, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.100730895996094, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8798222541809082, "num_tokens": 765115847.0, "step": 20051 }, { "epoch": 2.5508205062969087, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.180397987365723, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8725674152374268, "num_tokens": 765146839.0, "step": 20052 }, { "epoch": 2.5509477165754992, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.097853660583496, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.873519241809845, "num_tokens": 765185921.0, "step": 20053 }, { "epoch": 2.5510749268540898, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.150633811950684, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8736417293548584, "num_tokens": 765224534.0, "step": 20054 }, { "epoch": 2.5512021371326803, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.028715133666992, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8814558386802673, "num_tokens": 765267160.0, "step": 20055 }, { "epoch": 2.551329347411271, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.162300109863281, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.874169111251831, "num_tokens": 765304770.0, "step": 20056 }, { "epoch": 2.5514565576898613, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.151841163635254, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.870425820350647, "num_tokens": 765341966.0, "step": 20057 }, { "epoch": 2.551583767968452, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.140427589416504, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8884405493736267, "num_tokens": 765382653.0, "step": 20058 }, { "epoch": 2.5517109782470424, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.167379379272461, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8714505434036255, "num_tokens": 765414940.0, "step": 20059 }, { "epoch": 2.551838188525633, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.174139976501465, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8759011030197144, "num_tokens": 765445549.0, "step": 20060 }, { "epoch": 2.5519653988042235, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.13885498046875, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8801108598709106, "num_tokens": 765484474.0, "step": 20061 }, { "epoch": 2.552092609082814, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.095468521118164, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8733842968940735, "num_tokens": 765522651.0, "step": 20062 }, { "epoch": 2.5522198193614045, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.139424324035645, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8682911396026611, "num_tokens": 765565983.0, "step": 20063 }, { "epoch": 2.552347029639995, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.1610746383667, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8743911981582642, "num_tokens": 765604056.0, "step": 20064 }, { "epoch": 2.5524742399185856, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.100136756896973, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8820775747299194, "num_tokens": 765646062.0, "step": 20065 }, { "epoch": 2.5526014501971757, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.285273551940918, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8724923133850098, "num_tokens": 765687284.0, "step": 20066 }, { "epoch": 2.5527286604757666, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.105114936828613, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8608583807945251, "num_tokens": 765723743.0, "step": 20067 }, { "epoch": 2.5528558707543567, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.184086799621582, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8697333931922913, "num_tokens": 765768472.0, "step": 20068 }, { "epoch": 2.5529830810329477, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.104273796081543, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8885840177536011, "num_tokens": 765809040.0, "step": 20069 }, { "epoch": 2.5531102913115378, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.181671142578125, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8744852542877197, "num_tokens": 765845968.0, "step": 20070 }, { "epoch": 2.5532375015901287, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.198088645935059, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8782551288604736, "num_tokens": 765881037.0, "step": 20071 }, { "epoch": 2.553364711868719, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.139243125915527, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8784623146057129, "num_tokens": 765918945.0, "step": 20072 }, { "epoch": 2.5534919221473094, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.201562881469727, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.8859380483627319, "num_tokens": 765956128.0, "step": 20073 }, { "epoch": 2.5536191324259, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.144869804382324, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8861707448959351, "num_tokens": 765997109.0, "step": 20074 }, { "epoch": 2.5537463427044904, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.16446304321289, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8715969324111938, "num_tokens": 766032557.0, "step": 20075 }, { "epoch": 2.553873552983081, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.282249450683594, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8695294857025146, "num_tokens": 766075285.0, "step": 20076 }, { "epoch": 2.5540007632616715, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.098642349243164, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8716437816619873, "num_tokens": 766110600.0, "step": 20077 }, { "epoch": 2.554127973540262, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.177159309387207, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8629077672958374, "num_tokens": 766146877.0, "step": 20078 }, { "epoch": 2.5542551838188525, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.234207153320312, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8692145347595215, "num_tokens": 766185164.0, "step": 20079 }, { "epoch": 2.554382394097443, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.098551750183105, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8918927907943726, "num_tokens": 766221442.0, "step": 20080 }, { "epoch": 2.5545096043760336, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.210527420043945, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.850317120552063, "num_tokens": 766255258.0, "step": 20081 }, { "epoch": 2.554636814654624, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.165804862976074, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8710017204284668, "num_tokens": 766292392.0, "step": 20082 }, { "epoch": 2.5547640249332146, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.10713005065918, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8637375831604004, "num_tokens": 766335312.0, "step": 20083 }, { "epoch": 2.554891235211805, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.145846366882324, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8746380805969238, "num_tokens": 766372189.0, "step": 20084 }, { "epoch": 2.5550184454903957, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.165059089660645, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8631235361099243, "num_tokens": 766414021.0, "step": 20085 }, { "epoch": 2.5551456557689862, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.170368194580078, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8860452175140381, "num_tokens": 766452005.0, "step": 20086 }, { "epoch": 2.5552728660475768, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.214699745178223, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8682054281234741, "num_tokens": 766492150.0, "step": 20087 }, { "epoch": 2.5554000763261673, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.19335651397705, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8839008808135986, "num_tokens": 766533889.0, "step": 20088 }, { "epoch": 2.555527286604758, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.213643074035645, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8746107816696167, "num_tokens": 766565008.0, "step": 20089 }, { "epoch": 2.5556544968833483, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.175029754638672, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8744301795959473, "num_tokens": 766605744.0, "step": 20090 }, { "epoch": 2.5557817071619384, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.147244453430176, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8818156719207764, "num_tokens": 766645782.0, "step": 20091 }, { "epoch": 2.5559089174405294, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.179287910461426, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.864077091217041, "num_tokens": 766683135.0, "step": 20092 }, { "epoch": 2.5560361277191195, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.11639404296875, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8738865852355957, "num_tokens": 766720691.0, "step": 20093 }, { "epoch": 2.5561633379977104, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.18906307220459, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8686192035675049, "num_tokens": 766755426.0, "step": 20094 }, { "epoch": 2.5562905482763005, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.158710479736328, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8789948225021362, "num_tokens": 766788315.0, "step": 20095 }, { "epoch": 2.5564177585548915, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.096647262573242, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8896164894104004, "num_tokens": 766825664.0, "step": 20096 }, { "epoch": 2.5565449688334816, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.190842628479004, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8851655721664429, "num_tokens": 766868145.0, "step": 20097 }, { "epoch": 2.556672179112072, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.159987449645996, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.862372875213623, "num_tokens": 766907278.0, "step": 20098 }, { "epoch": 2.5567993893906626, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.199052810668945, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.886147677898407, "num_tokens": 766941662.0, "step": 20099 }, { "epoch": 2.556926599669253, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.203462600708008, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8559073805809021, "num_tokens": 766977725.0, "step": 20100 }, { "epoch": 2.5570538099478437, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.169069290161133, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8690786957740784, "num_tokens": 767011011.0, "step": 20101 }, { "epoch": 2.5571810202264342, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.171208381652832, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8712416291236877, "num_tokens": 767047078.0, "step": 20102 }, { "epoch": 2.5573082305050248, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.205060958862305, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8692000508308411, "num_tokens": 767089208.0, "step": 20103 }, { "epoch": 2.5574354407836153, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.174216270446777, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8579596281051636, "num_tokens": 767131225.0, "step": 20104 }, { "epoch": 2.557562651062206, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.164074897766113, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.862496018409729, "num_tokens": 767165222.0, "step": 20105 }, { "epoch": 2.5576898613407963, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.24987506866455, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8744076490402222, "num_tokens": 767200870.0, "step": 20106 }, { "epoch": 2.557817071619387, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.175155639648438, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8723587989807129, "num_tokens": 767236881.0, "step": 20107 }, { "epoch": 2.5579442818979774, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.214374542236328, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8787059783935547, "num_tokens": 767269716.0, "step": 20108 }, { "epoch": 2.558071492176568, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.256052017211914, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.864336371421814, "num_tokens": 767306840.0, "step": 20109 }, { "epoch": 2.5581987024551585, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.183351516723633, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8859676122665405, "num_tokens": 767346966.0, "step": 20110 }, { "epoch": 2.558325912733749, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.188206672668457, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8695436716079712, "num_tokens": 767387084.0, "step": 20111 }, { "epoch": 2.5584531230123395, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.271100044250488, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8539817333221436, "num_tokens": 767417857.0, "step": 20112 }, { "epoch": 2.55858033329093, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.21647834777832, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8749910593032837, "num_tokens": 767455771.0, "step": 20113 }, { "epoch": 2.5587075435695206, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.203709602355957, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8601043224334717, "num_tokens": 767493187.0, "step": 20114 }, { "epoch": 2.558834753848111, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.216215133666992, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8655179738998413, "num_tokens": 767534526.0, "step": 20115 }, { "epoch": 2.558961964126701, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.213664054870605, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8710737228393555, "num_tokens": 767569297.0, "step": 20116 }, { "epoch": 2.559089174405292, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.221895217895508, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8696355819702148, "num_tokens": 767605860.0, "step": 20117 }, { "epoch": 2.5592163846838822, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.189969062805176, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8758171796798706, "num_tokens": 767645252.0, "step": 20118 }, { "epoch": 2.559343594962473, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.253595352172852, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8738721609115601, "num_tokens": 767681942.0, "step": 20119 }, { "epoch": 2.5594708052410633, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.157803535461426, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8721448183059692, "num_tokens": 767720631.0, "step": 20120 }, { "epoch": 2.559598015519654, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.231521606445312, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8768918514251709, "num_tokens": 767757018.0, "step": 20121 }, { "epoch": 2.5597252257982444, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.178772926330566, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8771358132362366, "num_tokens": 767796046.0, "step": 20122 }, { "epoch": 2.559852436076835, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.162148475646973, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8832475543022156, "num_tokens": 767835678.0, "step": 20123 }, { "epoch": 2.5599796463554254, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.173821449279785, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8874781131744385, "num_tokens": 767870515.0, "step": 20124 }, { "epoch": 2.560106856634016, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.170063972473145, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8574692606925964, "num_tokens": 767911573.0, "step": 20125 }, { "epoch": 2.5602340669126065, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.165590286254883, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8722924590110779, "num_tokens": 767950161.0, "step": 20126 }, { "epoch": 2.560361277191197, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.143031120300293, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8718487024307251, "num_tokens": 767989431.0, "step": 20127 }, { "epoch": 2.5604884874697875, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.168322563171387, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8829044103622437, "num_tokens": 768023551.0, "step": 20128 }, { "epoch": 2.560615697748378, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.133868217468262, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8809536695480347, "num_tokens": 768055317.0, "step": 20129 }, { "epoch": 2.5607429080269686, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.115242004394531, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8669316172599792, "num_tokens": 768089575.0, "step": 20130 }, { "epoch": 2.560870118305559, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.15225887298584, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8640671968460083, "num_tokens": 768131795.0, "step": 20131 }, { "epoch": 2.5609973285841496, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.182005882263184, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8776441812515259, "num_tokens": 768165714.0, "step": 20132 }, { "epoch": 2.56112453886274, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.12459659576416, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8815293908119202, "num_tokens": 768203475.0, "step": 20133 }, { "epoch": 2.5612517491413307, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.218108177185059, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8892059922218323, "num_tokens": 768244841.0, "step": 20134 }, { "epoch": 2.561378959419921, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.201192855834961, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8683128952980042, "num_tokens": 768286502.0, "step": 20135 }, { "epoch": 2.5615061696985117, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.185100555419922, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8820608854293823, "num_tokens": 768320628.0, "step": 20136 }, { "epoch": 2.5616333799771023, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.225881576538086, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8669343590736389, "num_tokens": 768353630.0, "step": 20137 }, { "epoch": 2.561760590255693, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.187623023986816, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8730530142784119, "num_tokens": 768388170.0, "step": 20138 }, { "epoch": 2.5618878005342833, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.151982307434082, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8891478180885315, "num_tokens": 768428238.0, "step": 20139 }, { "epoch": 2.562015010812874, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.119873046875, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8872601985931396, "num_tokens": 768460075.0, "step": 20140 }, { "epoch": 2.562142221091464, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.198880195617676, "learning_rate": 1e-06, "loss": 0.339, "mean_token_accuracy": 0.8905069828033447, "num_tokens": 768493982.0, "step": 20141 }, { "epoch": 2.562269431370055, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.135696411132812, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8778987526893616, "num_tokens": 768529121.0, "step": 20142 }, { "epoch": 2.562396641648645, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.15945053100586, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8740462064743042, "num_tokens": 768560730.0, "step": 20143 }, { "epoch": 2.562523851927236, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.177505493164062, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8845638036727905, "num_tokens": 768603015.0, "step": 20144 }, { "epoch": 2.562651062205826, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.138669967651367, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8731200695037842, "num_tokens": 768639847.0, "step": 20145 }, { "epoch": 2.5627782724844166, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.177749633789062, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8709698915481567, "num_tokens": 768681614.0, "step": 20146 }, { "epoch": 2.562905482763007, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.228775978088379, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8772826194763184, "num_tokens": 768722052.0, "step": 20147 }, { "epoch": 2.5630326930415976, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.167325019836426, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8670347929000854, "num_tokens": 768754191.0, "step": 20148 }, { "epoch": 2.563159903320188, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.181830406188965, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.863756537437439, "num_tokens": 768792981.0, "step": 20149 }, { "epoch": 2.5632871135987787, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.186853408813477, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8636117577552795, "num_tokens": 768827040.0, "step": 20150 }, { "epoch": 2.5634143238773692, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.19247055053711, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8719662427902222, "num_tokens": 768865669.0, "step": 20151 }, { "epoch": 2.5635415341559598, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.158042907714844, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8729405999183655, "num_tokens": 768903306.0, "step": 20152 }, { "epoch": 2.5636687444345503, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.167281150817871, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8773099184036255, "num_tokens": 768946322.0, "step": 20153 }, { "epoch": 2.563795954713141, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.098297119140625, "learning_rate": 1e-06, "loss": 0.3389, "mean_token_accuracy": 0.8914403319358826, "num_tokens": 768987350.0, "step": 20154 }, { "epoch": 2.5639231649917313, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.169724464416504, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8828374743461609, "num_tokens": 769030762.0, "step": 20155 }, { "epoch": 2.564050375270322, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.1796293258667, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8844660520553589, "num_tokens": 769069827.0, "step": 20156 }, { "epoch": 2.5641775855489124, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.157950401306152, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8733208775520325, "num_tokens": 769107742.0, "step": 20157 }, { "epoch": 2.564304795827503, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.131134986877441, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8720040321350098, "num_tokens": 769153286.0, "step": 20158 }, { "epoch": 2.5644320061060935, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.132957458496094, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8754359483718872, "num_tokens": 769192646.0, "step": 20159 }, { "epoch": 2.564559216384684, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.167473793029785, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8519819974899292, "num_tokens": 769230360.0, "step": 20160 }, { "epoch": 2.5646864266632745, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.124754905700684, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.889671802520752, "num_tokens": 769270754.0, "step": 20161 }, { "epoch": 2.564813636941865, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.185222625732422, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8504446148872375, "num_tokens": 769310109.0, "step": 20162 }, { "epoch": 2.5649408472204556, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.135099411010742, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8725301027297974, "num_tokens": 769352101.0, "step": 20163 }, { "epoch": 2.5650680574990457, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.197184562683105, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8881096243858337, "num_tokens": 769386570.0, "step": 20164 }, { "epoch": 2.5651952677776366, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.187783241271973, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.884833812713623, "num_tokens": 769427167.0, "step": 20165 }, { "epoch": 2.5653224780562267, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.206975936889648, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8632495403289795, "num_tokens": 769465456.0, "step": 20166 }, { "epoch": 2.5654496883348177, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.18017864227295, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.886736273765564, "num_tokens": 769497135.0, "step": 20167 }, { "epoch": 2.5655768986134078, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.21937370300293, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8683546185493469, "num_tokens": 769533706.0, "step": 20168 }, { "epoch": 2.5657041088919987, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.15749740600586, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8683474659919739, "num_tokens": 769578698.0, "step": 20169 }, { "epoch": 2.565831319170589, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.217447280883789, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8859264850616455, "num_tokens": 769619799.0, "step": 20170 }, { "epoch": 2.5659585294491793, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.228501319885254, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8735621571540833, "num_tokens": 769655973.0, "step": 20171 }, { "epoch": 2.56608573972777, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.21496868133545, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8691220283508301, "num_tokens": 769692333.0, "step": 20172 }, { "epoch": 2.5662129500063604, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.15956974029541, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8779476284980774, "num_tokens": 769731804.0, "step": 20173 }, { "epoch": 2.566340160284951, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.262097358703613, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8795210123062134, "num_tokens": 769767409.0, "step": 20174 }, { "epoch": 2.5664673705635415, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.154633522033691, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8717100620269775, "num_tokens": 769806468.0, "step": 20175 }, { "epoch": 2.566594580842132, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.184216499328613, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8819273114204407, "num_tokens": 769842025.0, "step": 20176 }, { "epoch": 2.5667217911207225, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.184513092041016, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8689063787460327, "num_tokens": 769884125.0, "step": 20177 }, { "epoch": 2.566849001399313, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.171095848083496, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8633377552032471, "num_tokens": 769920523.0, "step": 20178 }, { "epoch": 2.5669762116779036, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.195384979248047, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8721712827682495, "num_tokens": 769960141.0, "step": 20179 }, { "epoch": 2.567103421956494, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.109659194946289, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8754477500915527, "num_tokens": 769995281.0, "step": 20180 }, { "epoch": 2.5672306322350846, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.246872901916504, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8775125741958618, "num_tokens": 770028230.0, "step": 20181 }, { "epoch": 2.567357842513675, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.139561653137207, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8802996873855591, "num_tokens": 770068996.0, "step": 20182 }, { "epoch": 2.5674850527922657, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 14.10732364654541, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8638924956321716, "num_tokens": 770107018.0, "step": 20183 }, { "epoch": 2.567612263070856, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.156806945800781, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8725225329399109, "num_tokens": 770140251.0, "step": 20184 }, { "epoch": 2.5677394733494467, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.151976585388184, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8661415576934814, "num_tokens": 770182241.0, "step": 20185 }, { "epoch": 2.5678666836280373, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.14235782623291, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8881077766418457, "num_tokens": 770223267.0, "step": 20186 }, { "epoch": 2.567993893906628, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.215252876281738, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8635565042495728, "num_tokens": 770261856.0, "step": 20187 }, { "epoch": 2.5681211041852183, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.129721641540527, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8635642528533936, "num_tokens": 770299762.0, "step": 20188 }, { "epoch": 2.5682483144638084, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.221434593200684, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8688139915466309, "num_tokens": 770333690.0, "step": 20189 }, { "epoch": 2.5683755247423994, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.135323524475098, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8802904486656189, "num_tokens": 770375755.0, "step": 20190 }, { "epoch": 2.5685027350209895, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.201641082763672, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8831287622451782, "num_tokens": 770417805.0, "step": 20191 }, { "epoch": 2.5686299452995804, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.191991806030273, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.872510552406311, "num_tokens": 770458183.0, "step": 20192 }, { "epoch": 2.5687571555781705, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.238621711730957, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8766450881958008, "num_tokens": 770488973.0, "step": 20193 }, { "epoch": 2.5688843658567615, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.120739936828613, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8680719137191772, "num_tokens": 770529030.0, "step": 20194 }, { "epoch": 2.5690115761353516, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.258273124694824, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8629112839698792, "num_tokens": 770574697.0, "step": 20195 }, { "epoch": 2.569138786413942, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.144038200378418, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8795135021209717, "num_tokens": 770615417.0, "step": 20196 }, { "epoch": 2.5692659966925326, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.165640830993652, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.860418438911438, "num_tokens": 770653228.0, "step": 20197 }, { "epoch": 2.569393206971123, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.241284370422363, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8884537816047668, "num_tokens": 770693876.0, "step": 20198 }, { "epoch": 2.5695204172497137, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.187976837158203, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8724650144577026, "num_tokens": 770728754.0, "step": 20199 }, { "epoch": 2.5696476275283042, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.225351333618164, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.871257483959198, "num_tokens": 770767961.0, "step": 20200 }, { "epoch": 2.5697748378068948, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.188263893127441, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8466099500656128, "num_tokens": 770808880.0, "step": 20201 }, { "epoch": 2.5699020480854853, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.206947326660156, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.868061900138855, "num_tokens": 770846483.0, "step": 20202 }, { "epoch": 2.570029258364076, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.190924644470215, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8539548516273499, "num_tokens": 770884234.0, "step": 20203 }, { "epoch": 2.5701564686426663, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.180547714233398, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8595713376998901, "num_tokens": 770918785.0, "step": 20204 }, { "epoch": 2.570283678921257, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.17923355102539, "learning_rate": 1e-06, "loss": 0.3342, "mean_token_accuracy": 0.8950765132904053, "num_tokens": 770956889.0, "step": 20205 }, { "epoch": 2.5704108891998474, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.272278785705566, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.860074520111084, "num_tokens": 770994937.0, "step": 20206 }, { "epoch": 2.570538099478438, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.193192481994629, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8822087645530701, "num_tokens": 771031451.0, "step": 20207 }, { "epoch": 2.5706653097570284, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.215766906738281, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8810893297195435, "num_tokens": 771061575.0, "step": 20208 }, { "epoch": 2.570792520035619, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.210437774658203, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8718223571777344, "num_tokens": 771103366.0, "step": 20209 }, { "epoch": 2.5709197303142095, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.162232398986816, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8775132298469543, "num_tokens": 771142178.0, "step": 20210 }, { "epoch": 2.5710469405928, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.296971321105957, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8740118741989136, "num_tokens": 771179296.0, "step": 20211 }, { "epoch": 2.5711741508713906, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.18790340423584, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8668714165687561, "num_tokens": 771219084.0, "step": 20212 }, { "epoch": 2.571301361149981, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.206271171569824, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8803567290306091, "num_tokens": 771250609.0, "step": 20213 }, { "epoch": 2.571428571428571, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.218960762023926, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8718710541725159, "num_tokens": 771286191.0, "step": 20214 }, { "epoch": 2.571555781707162, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.158263206481934, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.87232905626297, "num_tokens": 771320898.0, "step": 20215 }, { "epoch": 2.5716829919857522, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.240458488464355, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.864778459072113, "num_tokens": 771353977.0, "step": 20216 }, { "epoch": 2.571810202264343, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.192599296569824, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8784788846969604, "num_tokens": 771391960.0, "step": 20217 }, { "epoch": 2.5719374125429333, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.121576309204102, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.865888774394989, "num_tokens": 771427764.0, "step": 20218 }, { "epoch": 2.572064622821524, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.235199928283691, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.873146653175354, "num_tokens": 771469706.0, "step": 20219 }, { "epoch": 2.5721918331001143, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.127616882324219, "learning_rate": 1e-06, "loss": 0.3341, "mean_token_accuracy": 0.8941388130187988, "num_tokens": 771504296.0, "step": 20220 }, { "epoch": 2.572319043378705, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.236948013305664, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8736904859542847, "num_tokens": 771544496.0, "step": 20221 }, { "epoch": 2.5724462536572954, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.157108306884766, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8823144435882568, "num_tokens": 771584322.0, "step": 20222 }, { "epoch": 2.572573463935886, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.123867988586426, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8823583126068115, "num_tokens": 771621424.0, "step": 20223 }, { "epoch": 2.5727006742144765, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.182236671447754, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.875079333782196, "num_tokens": 771658173.0, "step": 20224 }, { "epoch": 2.572827884493067, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.208654403686523, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8653469085693359, "num_tokens": 771698182.0, "step": 20225 }, { "epoch": 2.5729550947716575, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.265453338623047, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8572503924369812, "num_tokens": 771731559.0, "step": 20226 }, { "epoch": 2.573082305050248, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.13502311706543, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8693296313285828, "num_tokens": 771774362.0, "step": 20227 }, { "epoch": 2.5732095153288386, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.206228256225586, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8632441759109497, "num_tokens": 771810540.0, "step": 20228 }, { "epoch": 2.573336725607429, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.246695518493652, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8774453997612, "num_tokens": 771846221.0, "step": 20229 }, { "epoch": 2.5734639358860196, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.170654296875, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8828673958778381, "num_tokens": 771882268.0, "step": 20230 }, { "epoch": 2.57359114616461, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.326616287231445, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8727496862411499, "num_tokens": 771913176.0, "step": 20231 }, { "epoch": 2.5737183564432007, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.197617530822754, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8686365485191345, "num_tokens": 771948163.0, "step": 20232 }, { "epoch": 2.573845566721791, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.189532279968262, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8861592411994934, "num_tokens": 771988296.0, "step": 20233 }, { "epoch": 2.5739727770003817, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.15170669555664, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8810974955558777, "num_tokens": 772024330.0, "step": 20234 }, { "epoch": 2.5740999872789723, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.263457298278809, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8613647222518921, "num_tokens": 772062291.0, "step": 20235 }, { "epoch": 2.574227197557563, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.10600757598877, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.874648928642273, "num_tokens": 772110239.0, "step": 20236 }, { "epoch": 2.5743544078361533, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.20348072052002, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.869064211845398, "num_tokens": 772153201.0, "step": 20237 }, { "epoch": 2.574481618114744, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.206258773803711, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8754220008850098, "num_tokens": 772186816.0, "step": 20238 }, { "epoch": 2.574608828393334, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.161043167114258, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8831109404563904, "num_tokens": 772224238.0, "step": 20239 }, { "epoch": 2.574736038671925, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.181838035583496, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8819370865821838, "num_tokens": 772263505.0, "step": 20240 }, { "epoch": 2.574863248950515, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.181824684143066, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.881826639175415, "num_tokens": 772303205.0, "step": 20241 }, { "epoch": 2.574990459229106, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.2153959274292, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8787944316864014, "num_tokens": 772344495.0, "step": 20242 }, { "epoch": 2.575117669507696, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.176178932189941, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8685159087181091, "num_tokens": 772380296.0, "step": 20243 }, { "epoch": 2.5752448797862866, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.216118812561035, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8864802718162537, "num_tokens": 772423933.0, "step": 20244 }, { "epoch": 2.575372090064877, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.129193305969238, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.860495924949646, "num_tokens": 772464514.0, "step": 20245 }, { "epoch": 2.5754993003434676, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.21036434173584, "learning_rate": 1e-06, "loss": 0.3345, "mean_token_accuracy": 0.8921374082565308, "num_tokens": 772498501.0, "step": 20246 }, { "epoch": 2.575626510622058, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.218073844909668, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.869780421257019, "num_tokens": 772538072.0, "step": 20247 }, { "epoch": 2.5757537209006487, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.167394638061523, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8739103078842163, "num_tokens": 772577434.0, "step": 20248 }, { "epoch": 2.575880931179239, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.217830657958984, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8875824809074402, "num_tokens": 772612917.0, "step": 20249 }, { "epoch": 2.5760081414578297, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.164117813110352, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8849236369132996, "num_tokens": 772652939.0, "step": 20250 }, { "epoch": 2.5761353517364203, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.192235946655273, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8802855014801025, "num_tokens": 772688018.0, "step": 20251 }, { "epoch": 2.576262562015011, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.198124885559082, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8643450736999512, "num_tokens": 772728071.0, "step": 20252 }, { "epoch": 2.5763897722936013, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.321255683898926, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8755542635917664, "num_tokens": 772762730.0, "step": 20253 }, { "epoch": 2.576516982572192, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.212508201599121, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8768129348754883, "num_tokens": 772799125.0, "step": 20254 }, { "epoch": 2.5766441928507824, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.17652702331543, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8861057758331299, "num_tokens": 772838657.0, "step": 20255 }, { "epoch": 2.576771403129373, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.176499366760254, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8818854093551636, "num_tokens": 772876470.0, "step": 20256 }, { "epoch": 2.5768986134079634, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.201859474182129, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.867846667766571, "num_tokens": 772910369.0, "step": 20257 }, { "epoch": 2.577025823686554, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.226408004760742, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8790024518966675, "num_tokens": 772948129.0, "step": 20258 }, { "epoch": 2.5771530339651445, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.25236701965332, "learning_rate": 1e-06, "loss": 0.3192, "mean_token_accuracy": 0.9009945392608643, "num_tokens": 772985359.0, "step": 20259 }, { "epoch": 2.577280244243735, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.227867126464844, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8755183815956116, "num_tokens": 773019402.0, "step": 20260 }, { "epoch": 2.5774074545223256, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.231968879699707, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8731728196144104, "num_tokens": 773059176.0, "step": 20261 }, { "epoch": 2.5775346648009156, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.257285118103027, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8764832615852356, "num_tokens": 773094274.0, "step": 20262 }, { "epoch": 2.5776618750795066, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.204029083251953, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8733025789260864, "num_tokens": 773135679.0, "step": 20263 }, { "epoch": 2.5777890853580967, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.20984935760498, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.870607316493988, "num_tokens": 773174270.0, "step": 20264 }, { "epoch": 2.5779162956366877, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.162839889526367, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8759574294090271, "num_tokens": 773213582.0, "step": 20265 }, { "epoch": 2.5780435059152778, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.294407844543457, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8690659403800964, "num_tokens": 773244980.0, "step": 20266 }, { "epoch": 2.5781707161938687, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.178972244262695, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8884470462799072, "num_tokens": 773287395.0, "step": 20267 }, { "epoch": 2.578297926472459, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.273032188415527, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8708808422088623, "num_tokens": 773324593.0, "step": 20268 }, { "epoch": 2.5784251367510493, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.185630798339844, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8693804740905762, "num_tokens": 773361802.0, "step": 20269 }, { "epoch": 2.57855234702964, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.230448722839355, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8792010545730591, "num_tokens": 773401348.0, "step": 20270 }, { "epoch": 2.5786795573082304, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.226475715637207, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8903440237045288, "num_tokens": 773439040.0, "step": 20271 }, { "epoch": 2.578806767586821, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.261605262756348, "learning_rate": 1e-06, "loss": 0.3491, "mean_token_accuracy": 0.8859826326370239, "num_tokens": 773471382.0, "step": 20272 }, { "epoch": 2.5789339778654115, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.20841121673584, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8634878396987915, "num_tokens": 773508924.0, "step": 20273 }, { "epoch": 2.579061188144002, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.217266082763672, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8887341618537903, "num_tokens": 773550219.0, "step": 20274 }, { "epoch": 2.5791883984225925, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.152836799621582, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8695932626724243, "num_tokens": 773589363.0, "step": 20275 }, { "epoch": 2.579315608701183, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.26492691040039, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8732556104660034, "num_tokens": 773630289.0, "step": 20276 }, { "epoch": 2.5794428189797736, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.21255111694336, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8755261301994324, "num_tokens": 773668660.0, "step": 20277 }, { "epoch": 2.579570029258364, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.22382926940918, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8713468909263611, "num_tokens": 773704318.0, "step": 20278 }, { "epoch": 2.5796972395369546, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.316946983337402, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8597699403762817, "num_tokens": 773745821.0, "step": 20279 }, { "epoch": 2.579824449815545, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.20906925201416, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8831682801246643, "num_tokens": 773782880.0, "step": 20280 }, { "epoch": 2.5799516600941357, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.24769401550293, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8609200716018677, "num_tokens": 773824880.0, "step": 20281 }, { "epoch": 2.580078870372726, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.250692367553711, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8777742385864258, "num_tokens": 773863210.0, "step": 20282 }, { "epoch": 2.5802060806513167, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.208060264587402, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8808132410049438, "num_tokens": 773903432.0, "step": 20283 }, { "epoch": 2.5803332909299073, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.174734115600586, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8819692134857178, "num_tokens": 773942424.0, "step": 20284 }, { "epoch": 2.580460501208498, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.198265075683594, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8741018176078796, "num_tokens": 773976598.0, "step": 20285 }, { "epoch": 2.5805877114870883, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.160797119140625, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8681395649909973, "num_tokens": 774016696.0, "step": 20286 }, { "epoch": 2.5807149217656784, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.184178352355957, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.868248462677002, "num_tokens": 774059046.0, "step": 20287 }, { "epoch": 2.5808421320442694, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.115784645080566, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8702279925346375, "num_tokens": 774095322.0, "step": 20288 }, { "epoch": 2.5809693423228595, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.25969123840332, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8627282381057739, "num_tokens": 774134506.0, "step": 20289 }, { "epoch": 2.5810965526014504, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.147241592407227, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8888362646102905, "num_tokens": 774176997.0, "step": 20290 }, { "epoch": 2.5812237628800405, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.250207901000977, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8695907592773438, "num_tokens": 774213900.0, "step": 20291 }, { "epoch": 2.5813509731586315, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.222752571105957, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8663179278373718, "num_tokens": 774259687.0, "step": 20292 }, { "epoch": 2.5814781834372216, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.123284339904785, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8784849643707275, "num_tokens": 774297200.0, "step": 20293 }, { "epoch": 2.581605393715812, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.263705253601074, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8865674734115601, "num_tokens": 774336642.0, "step": 20294 }, { "epoch": 2.5817326039944026, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.18896198272705, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8586031198501587, "num_tokens": 774376762.0, "step": 20295 }, { "epoch": 2.581859814272993, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.146949768066406, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8823226094245911, "num_tokens": 774421360.0, "step": 20296 }, { "epoch": 2.5819870245515837, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.243317604064941, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8813896775245667, "num_tokens": 774460850.0, "step": 20297 }, { "epoch": 2.582114234830174, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.251119613647461, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8712774515151978, "num_tokens": 774497542.0, "step": 20298 }, { "epoch": 2.5822414451087647, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.248320579528809, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8658633232116699, "num_tokens": 774538164.0, "step": 20299 }, { "epoch": 2.5823686553873553, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.183053970336914, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8685119152069092, "num_tokens": 774576118.0, "step": 20300 }, { "epoch": 2.582495865665946, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.180012702941895, "learning_rate": 1e-06, "loss": 0.3369, "mean_token_accuracy": 0.8954394459724426, "num_tokens": 774617506.0, "step": 20301 }, { "epoch": 2.5826230759445363, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.275989532470703, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8505817651748657, "num_tokens": 774653720.0, "step": 20302 }, { "epoch": 2.582750286223127, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.248615264892578, "learning_rate": 1e-06, "loss": 0.3307, "mean_token_accuracy": 0.8948028087615967, "num_tokens": 774691056.0, "step": 20303 }, { "epoch": 2.5828774965017174, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.225639343261719, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8684697151184082, "num_tokens": 774726514.0, "step": 20304 }, { "epoch": 2.583004706780308, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.295557975769043, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8660842180252075, "num_tokens": 774760831.0, "step": 20305 }, { "epoch": 2.5831319170588984, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.168867111206055, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8841565847396851, "num_tokens": 774799399.0, "step": 20306 }, { "epoch": 2.583259127337489, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.243755340576172, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8662599325180054, "num_tokens": 774839910.0, "step": 20307 }, { "epoch": 2.5833863376160795, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.233882904052734, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8748339414596558, "num_tokens": 774875694.0, "step": 20308 }, { "epoch": 2.58351354789467, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.242410659790039, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8762444257736206, "num_tokens": 774908020.0, "step": 20309 }, { "epoch": 2.5836407581732606, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.267254829406738, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8642701506614685, "num_tokens": 774946716.0, "step": 20310 }, { "epoch": 2.583767968451851, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.217832565307617, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.868590772151947, "num_tokens": 774981732.0, "step": 20311 }, { "epoch": 2.583895178730441, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.197077751159668, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8797857761383057, "num_tokens": 775015280.0, "step": 20312 }, { "epoch": 2.584022389009032, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.182394027709961, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8814237117767334, "num_tokens": 775054518.0, "step": 20313 }, { "epoch": 2.5841495992876222, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.215033531188965, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8802723288536072, "num_tokens": 775091548.0, "step": 20314 }, { "epoch": 2.584276809566213, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.247759819030762, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8622110486030579, "num_tokens": 775130916.0, "step": 20315 }, { "epoch": 2.5844040198448033, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.241172790527344, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8860653638839722, "num_tokens": 775165875.0, "step": 20316 }, { "epoch": 2.584531230123394, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.226350784301758, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8607305288314819, "num_tokens": 775211463.0, "step": 20317 }, { "epoch": 2.5846584404019843, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.218762397766113, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.882358193397522, "num_tokens": 775248338.0, "step": 20318 }, { "epoch": 2.584785650680575, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.236204147338867, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8631254434585571, "num_tokens": 775285707.0, "step": 20319 }, { "epoch": 2.5849128609591654, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.235063552856445, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8828245401382446, "num_tokens": 775322682.0, "step": 20320 }, { "epoch": 2.585040071237756, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.279656410217285, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8838227987289429, "num_tokens": 775358274.0, "step": 20321 }, { "epoch": 2.5851672815163464, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.196626663208008, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8695970773696899, "num_tokens": 775394472.0, "step": 20322 }, { "epoch": 2.585294491794937, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.2415771484375, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8845125436782837, "num_tokens": 775433647.0, "step": 20323 }, { "epoch": 2.5854217020735275, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.262202262878418, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8653954267501831, "num_tokens": 775469159.0, "step": 20324 }, { "epoch": 2.585548912352118, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.233115196228027, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8823703527450562, "num_tokens": 775507329.0, "step": 20325 }, { "epoch": 2.5856761226307086, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.172533988952637, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8720342516899109, "num_tokens": 775544638.0, "step": 20326 }, { "epoch": 2.585803332909299, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.244300842285156, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.876401424407959, "num_tokens": 775584918.0, "step": 20327 }, { "epoch": 2.5859305431878896, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.155792236328125, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8696811199188232, "num_tokens": 775624214.0, "step": 20328 }, { "epoch": 2.58605775346648, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.281028747558594, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8621923923492432, "num_tokens": 775668372.0, "step": 20329 }, { "epoch": 2.5861849637450707, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.225975036621094, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.846497118473053, "num_tokens": 775700578.0, "step": 20330 }, { "epoch": 2.586312174023661, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.2192964553833, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8798482418060303, "num_tokens": 775746980.0, "step": 20331 }, { "epoch": 2.5864393843022517, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.210902214050293, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8681687712669373, "num_tokens": 775780762.0, "step": 20332 }, { "epoch": 2.5865665945808423, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.190899848937988, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8802012801170349, "num_tokens": 775815591.0, "step": 20333 }, { "epoch": 2.586693804859433, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.262482643127441, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8656030893325806, "num_tokens": 775856414.0, "step": 20334 }, { "epoch": 2.5868210151380233, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.19559097290039, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.874639630317688, "num_tokens": 775896939.0, "step": 20335 }, { "epoch": 2.586948225416614, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.201284408569336, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8684966564178467, "num_tokens": 775941778.0, "step": 20336 }, { "epoch": 2.587075435695204, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.193361282348633, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8744073510169983, "num_tokens": 775979346.0, "step": 20337 }, { "epoch": 2.587202645973795, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.253058433532715, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8722171187400818, "num_tokens": 776013793.0, "step": 20338 }, { "epoch": 2.587329856252385, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.216196060180664, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8857700824737549, "num_tokens": 776058513.0, "step": 20339 }, { "epoch": 2.587457066530976, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.189911842346191, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8746010661125183, "num_tokens": 776102790.0, "step": 20340 }, { "epoch": 2.587584276809566, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.296353340148926, "learning_rate": 1e-06, "loss": 0.347, "mean_token_accuracy": 0.8893876075744629, "num_tokens": 776133898.0, "step": 20341 }, { "epoch": 2.5877114870881566, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.202217102050781, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8719617128372192, "num_tokens": 776167590.0, "step": 20342 }, { "epoch": 2.587838697366747, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.223740577697754, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.868878185749054, "num_tokens": 776207966.0, "step": 20343 }, { "epoch": 2.5879659076453376, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.19654369354248, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8745665550231934, "num_tokens": 776245601.0, "step": 20344 }, { "epoch": 2.588093117923928, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.237409591674805, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.883110761642456, "num_tokens": 776285377.0, "step": 20345 }, { "epoch": 2.5882203282025187, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2178373336792, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8651543855667114, "num_tokens": 776328116.0, "step": 20346 }, { "epoch": 2.588347538481109, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.178118705749512, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8589829802513123, "num_tokens": 776369008.0, "step": 20347 }, { "epoch": 2.5884747487596997, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.191131591796875, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8651653528213501, "num_tokens": 776410151.0, "step": 20348 }, { "epoch": 2.5886019590382903, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.179580688476562, "learning_rate": 1e-06, "loss": 0.3397, "mean_token_accuracy": 0.898402988910675, "num_tokens": 776445944.0, "step": 20349 }, { "epoch": 2.588729169316881, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.212169647216797, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8778574466705322, "num_tokens": 776485388.0, "step": 20350 }, { "epoch": 2.5888563795954713, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.230018615722656, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8565061688423157, "num_tokens": 776529135.0, "step": 20351 }, { "epoch": 2.588983589874062, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.204305648803711, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8674224615097046, "num_tokens": 776564310.0, "step": 20352 }, { "epoch": 2.5891108001526524, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.255682945251465, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8871139287948608, "num_tokens": 776609004.0, "step": 20353 }, { "epoch": 2.589238010431243, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.17652416229248, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8801162242889404, "num_tokens": 776645796.0, "step": 20354 }, { "epoch": 2.5893652207098334, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.209793090820312, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8653863668441772, "num_tokens": 776684248.0, "step": 20355 }, { "epoch": 2.589492430988424, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.203174591064453, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8773981332778931, "num_tokens": 776722684.0, "step": 20356 }, { "epoch": 2.5896196412670145, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.192442893981934, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8784728646278381, "num_tokens": 776764635.0, "step": 20357 }, { "epoch": 2.589746851545605, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.24587345123291, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.863100528717041, "num_tokens": 776805683.0, "step": 20358 }, { "epoch": 2.5898740618241956, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.182024002075195, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8771045207977295, "num_tokens": 776841229.0, "step": 20359 }, { "epoch": 2.5900012721027856, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.208860397338867, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.880884051322937, "num_tokens": 776881319.0, "step": 20360 }, { "epoch": 2.5901284823813766, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.26928424835205, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8789234161376953, "num_tokens": 776919436.0, "step": 20361 }, { "epoch": 2.5902556926599667, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.185781478881836, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8559980392456055, "num_tokens": 776958644.0, "step": 20362 }, { "epoch": 2.5903829029385577, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.273374557495117, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.8922690153121948, "num_tokens": 776994509.0, "step": 20363 }, { "epoch": 2.5905101132171477, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.233997344970703, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8798168897628784, "num_tokens": 777037557.0, "step": 20364 }, { "epoch": 2.5906373234957387, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.211591720581055, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8544380068778992, "num_tokens": 777077237.0, "step": 20365 }, { "epoch": 2.590764533774329, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.248319625854492, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8811353445053101, "num_tokens": 777120813.0, "step": 20366 }, { "epoch": 2.5908917440529193, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.2074613571167, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.861953854560852, "num_tokens": 777155809.0, "step": 20367 }, { "epoch": 2.59101895433151, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2919921875, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.874541163444519, "num_tokens": 777195868.0, "step": 20368 }, { "epoch": 2.5911461646101004, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.193252563476562, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8784410357475281, "num_tokens": 777233406.0, "step": 20369 }, { "epoch": 2.591273374888691, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.249317169189453, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8754614591598511, "num_tokens": 777272817.0, "step": 20370 }, { "epoch": 2.5914005851672814, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.269137382507324, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.870479941368103, "num_tokens": 777312398.0, "step": 20371 }, { "epoch": 2.591527795445872, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.180953979492188, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8847652673721313, "num_tokens": 777349375.0, "step": 20372 }, { "epoch": 2.5916550057244625, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.253581047058105, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.88396817445755, "num_tokens": 777394806.0, "step": 20373 }, { "epoch": 2.591782216003053, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.13918685913086, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8783888816833496, "num_tokens": 777434350.0, "step": 20374 }, { "epoch": 2.5919094262816436, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.301048278808594, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8486931324005127, "num_tokens": 777469690.0, "step": 20375 }, { "epoch": 2.592036636560234, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.251781463623047, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8675940036773682, "num_tokens": 777502095.0, "step": 20376 }, { "epoch": 2.5921638468388246, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.250594139099121, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8857556581497192, "num_tokens": 777540387.0, "step": 20377 }, { "epoch": 2.592291057117415, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2482328414917, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8812214732170105, "num_tokens": 777581855.0, "step": 20378 }, { "epoch": 2.5924182673960057, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.246355056762695, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8757181167602539, "num_tokens": 777619852.0, "step": 20379 }, { "epoch": 2.592545477674596, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.181888580322266, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8682519197463989, "num_tokens": 777662772.0, "step": 20380 }, { "epoch": 2.5926726879531867, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.251701354980469, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.857364296913147, "num_tokens": 777701171.0, "step": 20381 }, { "epoch": 2.5927998982317773, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.156954765319824, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8608146905899048, "num_tokens": 777738245.0, "step": 20382 }, { "epoch": 2.592927108510368, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.196557998657227, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8743699789047241, "num_tokens": 777775732.0, "step": 20383 }, { "epoch": 2.5930543187889583, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.263741493225098, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8748701810836792, "num_tokens": 777811077.0, "step": 20384 }, { "epoch": 2.5931815290675484, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.144229888916016, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8726078271865845, "num_tokens": 777854107.0, "step": 20385 }, { "epoch": 2.5933087393461394, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.265419960021973, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8728592991828918, "num_tokens": 777891950.0, "step": 20386 }, { "epoch": 2.5934359496247295, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.159978866577148, "learning_rate": 1e-06, "loss": 0.3418, "mean_token_accuracy": 0.8921993374824524, "num_tokens": 777932638.0, "step": 20387 }, { "epoch": 2.5935631599033204, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.274582862854004, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8725808262825012, "num_tokens": 777977733.0, "step": 20388 }, { "epoch": 2.5936903701819105, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.252756118774414, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8787593841552734, "num_tokens": 778020451.0, "step": 20389 }, { "epoch": 2.5938175804605015, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.178934097290039, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8800989985466003, "num_tokens": 778059594.0, "step": 20390 }, { "epoch": 2.5939447907390916, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.28994369506836, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8780515193939209, "num_tokens": 778099235.0, "step": 20391 }, { "epoch": 2.594072001017682, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.243562698364258, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8852806091308594, "num_tokens": 778135430.0, "step": 20392 }, { "epoch": 2.5941992112962726, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.308591842651367, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8675603866577148, "num_tokens": 778175111.0, "step": 20393 }, { "epoch": 2.594326421574863, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.227300643920898, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8650603294372559, "num_tokens": 778211954.0, "step": 20394 }, { "epoch": 2.5944536318534537, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.304929733276367, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8588623404502869, "num_tokens": 778246580.0, "step": 20395 }, { "epoch": 2.594580842132044, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.264473915100098, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8800802230834961, "num_tokens": 778288659.0, "step": 20396 }, { "epoch": 2.5947080524106347, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.255071640014648, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8731736540794373, "num_tokens": 778325365.0, "step": 20397 }, { "epoch": 2.5948352626892253, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.23569393157959, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.880810022354126, "num_tokens": 778362982.0, "step": 20398 }, { "epoch": 2.594962472967816, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.250579833984375, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8610588312149048, "num_tokens": 778394230.0, "step": 20399 }, { "epoch": 2.5950896832464063, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.286837577819824, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8637573719024658, "num_tokens": 778434345.0, "step": 20400 }, { "epoch": 2.595216893524997, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2847261428833, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8693709969520569, "num_tokens": 778478303.0, "step": 20401 }, { "epoch": 2.5953441038035874, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.295587539672852, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8804389834403992, "num_tokens": 778513915.0, "step": 20402 }, { "epoch": 2.595471314082178, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.216428756713867, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8900816440582275, "num_tokens": 778554294.0, "step": 20403 }, { "epoch": 2.5955985243607684, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.371020317077637, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8716500401496887, "num_tokens": 778592836.0, "step": 20404 }, { "epoch": 2.595725734639359, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.254958152770996, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8659390211105347, "num_tokens": 778629316.0, "step": 20405 }, { "epoch": 2.5958529449179495, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.325627326965332, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8710892796516418, "num_tokens": 778659065.0, "step": 20406 }, { "epoch": 2.59598015519654, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.1756010055542, "learning_rate": 1e-06, "loss": 0.3485, "mean_token_accuracy": 0.8900331258773804, "num_tokens": 778693949.0, "step": 20407 }, { "epoch": 2.5961073654751305, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.269434928894043, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8794248700141907, "num_tokens": 778738045.0, "step": 20408 }, { "epoch": 2.596234575753721, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.305105209350586, "learning_rate": 1e-06, "loss": 0.5188, "mean_token_accuracy": 0.8324723243713379, "num_tokens": 778769578.0, "step": 20409 }, { "epoch": 2.596361786032311, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.263799667358398, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.870136559009552, "num_tokens": 778810052.0, "step": 20410 }, { "epoch": 2.596488996310902, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.204878807067871, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8674688935279846, "num_tokens": 778850868.0, "step": 20411 }, { "epoch": 2.596616206589492, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.255940437316895, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8802286386489868, "num_tokens": 778890129.0, "step": 20412 }, { "epoch": 2.596743416868083, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.272369384765625, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8792445063591003, "num_tokens": 778923181.0, "step": 20413 }, { "epoch": 2.5968706271466733, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.169897079467773, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.866236686706543, "num_tokens": 778963328.0, "step": 20414 }, { "epoch": 2.596997837425264, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.24025821685791, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8846390247344971, "num_tokens": 779003932.0, "step": 20415 }, { "epoch": 2.5971250477038543, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.248974800109863, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8642699718475342, "num_tokens": 779043913.0, "step": 20416 }, { "epoch": 2.597252257982445, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.243544578552246, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8713862895965576, "num_tokens": 779077126.0, "step": 20417 }, { "epoch": 2.5973794682610354, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.254410743713379, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8780795931816101, "num_tokens": 779111544.0, "step": 20418 }, { "epoch": 2.597506678539626, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.233266830444336, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8730830550193787, "num_tokens": 779145493.0, "step": 20419 }, { "epoch": 2.5976338888182164, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.236019134521484, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8664743900299072, "num_tokens": 779189216.0, "step": 20420 }, { "epoch": 2.597761099096807, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.247857093811035, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8802380561828613, "num_tokens": 779225914.0, "step": 20421 }, { "epoch": 2.5978883093753975, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.283010482788086, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8781326413154602, "num_tokens": 779264118.0, "step": 20422 }, { "epoch": 2.598015519653988, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.298604011535645, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8766205310821533, "num_tokens": 779301694.0, "step": 20423 }, { "epoch": 2.5981427299325786, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.283330917358398, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8731809258460999, "num_tokens": 779339012.0, "step": 20424 }, { "epoch": 2.598269940211169, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.288948059082031, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8788455724716187, "num_tokens": 779376358.0, "step": 20425 }, { "epoch": 2.5983971504897596, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.233763694763184, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.872992753982544, "num_tokens": 779416344.0, "step": 20426 }, { "epoch": 2.59852436076835, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.201602935791016, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8639167547225952, "num_tokens": 779452189.0, "step": 20427 }, { "epoch": 2.5986515710469407, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.296760559082031, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8725130558013916, "num_tokens": 779493513.0, "step": 20428 }, { "epoch": 2.598778781325531, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.264362335205078, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.87968909740448, "num_tokens": 779533061.0, "step": 20429 }, { "epoch": 2.5989059916041217, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.305888175964355, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8666176795959473, "num_tokens": 779570159.0, "step": 20430 }, { "epoch": 2.5990332018827123, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.213578224182129, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8715381026268005, "num_tokens": 779609066.0, "step": 20431 }, { "epoch": 2.599160412161303, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.304825782775879, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.874139130115509, "num_tokens": 779651087.0, "step": 20432 }, { "epoch": 2.5992876224398933, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.21599292755127, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8873437643051147, "num_tokens": 779689531.0, "step": 20433 }, { "epoch": 2.599414832718484, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.373886108398438, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8686548471450806, "num_tokens": 779724062.0, "step": 20434 }, { "epoch": 2.599542042997074, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.201136589050293, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8613621592521667, "num_tokens": 779766029.0, "step": 20435 }, { "epoch": 2.599669253275665, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.23812484741211, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8581631183624268, "num_tokens": 779811065.0, "step": 20436 }, { "epoch": 2.599796463554255, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.27379322052002, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8607968688011169, "num_tokens": 779852442.0, "step": 20437 }, { "epoch": 2.599923673832846, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.210975646972656, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.878598690032959, "num_tokens": 779885993.0, "step": 20438 }, { "epoch": 2.600050884111436, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.275799751281738, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.877193033695221, "num_tokens": 779927791.0, "step": 20439 }, { "epoch": 2.6001780943900266, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.210172653198242, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8767247200012207, "num_tokens": 779961517.0, "step": 20440 }, { "epoch": 2.600305304668617, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.258877754211426, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8609775304794312, "num_tokens": 779992902.0, "step": 20441 }, { "epoch": 2.6004325149472076, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.219467163085938, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.873967170715332, "num_tokens": 780032849.0, "step": 20442 }, { "epoch": 2.600559725225798, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.22489070892334, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8859435319900513, "num_tokens": 780072563.0, "step": 20443 }, { "epoch": 2.6006869355043887, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.182635307312012, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8739121556282043, "num_tokens": 780115299.0, "step": 20444 }, { "epoch": 2.600814145782979, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.332350730895996, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8723171353340149, "num_tokens": 780155792.0, "step": 20445 }, { "epoch": 2.6009413560615697, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.185758590698242, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8694642186164856, "num_tokens": 780189040.0, "step": 20446 }, { "epoch": 2.6010685663401603, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.1983003616333, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8727643489837646, "num_tokens": 780229463.0, "step": 20447 }, { "epoch": 2.601195776618751, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.203680038452148, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8858253955841064, "num_tokens": 780265961.0, "step": 20448 }, { "epoch": 2.6013229868973413, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.253478050231934, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8687686920166016, "num_tokens": 780303220.0, "step": 20449 }, { "epoch": 2.601450197175932, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.218738555908203, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8701217174530029, "num_tokens": 780340456.0, "step": 20450 }, { "epoch": 2.6015774074545224, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.285369873046875, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8717447519302368, "num_tokens": 780381402.0, "step": 20451 }, { "epoch": 2.601704617733113, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.22740364074707, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8782883882522583, "num_tokens": 780422571.0, "step": 20452 }, { "epoch": 2.6018318280117034, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.244019508361816, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8692575693130493, "num_tokens": 780462447.0, "step": 20453 }, { "epoch": 2.601959038290294, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.253098487854004, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8729524612426758, "num_tokens": 780499183.0, "step": 20454 }, { "epoch": 2.6020862485688845, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.222811698913574, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8610036373138428, "num_tokens": 780539216.0, "step": 20455 }, { "epoch": 2.602213458847475, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.160213470458984, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8797144293785095, "num_tokens": 780578725.0, "step": 20456 }, { "epoch": 2.6023406691260655, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.316150665283203, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8743816614151001, "num_tokens": 780615175.0, "step": 20457 }, { "epoch": 2.6024678794046556, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.170330047607422, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8765262365341187, "num_tokens": 780651186.0, "step": 20458 }, { "epoch": 2.6025950896832466, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.232891082763672, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8658982515335083, "num_tokens": 780685550.0, "step": 20459 }, { "epoch": 2.6027222999618367, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.230093002319336, "learning_rate": 1e-06, "loss": 0.3313, "mean_token_accuracy": 0.8937879800796509, "num_tokens": 780725477.0, "step": 20460 }, { "epoch": 2.6028495102404277, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.264007568359375, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8760678768157959, "num_tokens": 780762481.0, "step": 20461 }, { "epoch": 2.6029767205190177, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.22427749633789, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8536550402641296, "num_tokens": 780800305.0, "step": 20462 }, { "epoch": 2.6031039307976087, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.302124977111816, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8752566576004028, "num_tokens": 780836271.0, "step": 20463 }, { "epoch": 2.603231141076199, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.200033187866211, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8850889801979065, "num_tokens": 780877363.0, "step": 20464 }, { "epoch": 2.6033583513547893, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.21545124053955, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8624313473701477, "num_tokens": 780922982.0, "step": 20465 }, { "epoch": 2.60348556163338, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.269696235656738, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8721612095832825, "num_tokens": 780957925.0, "step": 20466 }, { "epoch": 2.6036127719119704, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.327704429626465, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8549783229827881, "num_tokens": 780996922.0, "step": 20467 }, { "epoch": 2.603739982190561, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.167009353637695, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8564264178276062, "num_tokens": 781037209.0, "step": 20468 }, { "epoch": 2.6038671924691514, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.259493827819824, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.867204487323761, "num_tokens": 781078077.0, "step": 20469 }, { "epoch": 2.603994402747742, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.221057891845703, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8727750778198242, "num_tokens": 781122516.0, "step": 20470 }, { "epoch": 2.6041216130263325, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.339600563049316, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8658884167671204, "num_tokens": 781156290.0, "step": 20471 }, { "epoch": 2.604248823304923, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.23926067352295, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8808398842811584, "num_tokens": 781192769.0, "step": 20472 }, { "epoch": 2.6043760335835135, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.30812931060791, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8776671886444092, "num_tokens": 781227594.0, "step": 20473 }, { "epoch": 2.604503243862104, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.209311485290527, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8848018050193787, "num_tokens": 781259369.0, "step": 20474 }, { "epoch": 2.6046304541406946, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.170949935913086, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8731573820114136, "num_tokens": 781296788.0, "step": 20475 }, { "epoch": 2.604757664419285, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.260560989379883, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8754239082336426, "num_tokens": 781333103.0, "step": 20476 }, { "epoch": 2.6048848746978757, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.134572982788086, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8813923001289368, "num_tokens": 781372691.0, "step": 20477 }, { "epoch": 2.605012084976466, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.286358833312988, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8692313432693481, "num_tokens": 781410480.0, "step": 20478 }, { "epoch": 2.6051392952550567, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.185285568237305, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8643571734428406, "num_tokens": 781454110.0, "step": 20479 }, { "epoch": 2.6052665055336472, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.264938354492188, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8704622983932495, "num_tokens": 781494034.0, "step": 20480 }, { "epoch": 2.6053937158122378, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.178285598754883, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8709861040115356, "num_tokens": 781536225.0, "step": 20481 }, { "epoch": 2.6055209260908283, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.234606742858887, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8789785504341125, "num_tokens": 781573987.0, "step": 20482 }, { "epoch": 2.6056481363694184, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.291720390319824, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8774951100349426, "num_tokens": 781608987.0, "step": 20483 }, { "epoch": 2.6057753466480094, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.248564720153809, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8802137970924377, "num_tokens": 781646185.0, "step": 20484 }, { "epoch": 2.6059025569265994, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.263023376464844, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8632174730300903, "num_tokens": 781682999.0, "step": 20485 }, { "epoch": 2.6060297672051904, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.214567184448242, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8712022304534912, "num_tokens": 781726052.0, "step": 20486 }, { "epoch": 2.6061569774837805, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.272204399108887, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.887309730052948, "num_tokens": 781760562.0, "step": 20487 }, { "epoch": 2.6062841877623715, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.217172622680664, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8625840544700623, "num_tokens": 781803765.0, "step": 20488 }, { "epoch": 2.6064113980409616, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.29706859588623, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8747480511665344, "num_tokens": 781845419.0, "step": 20489 }, { "epoch": 2.606538608319552, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.202781677246094, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8822519779205322, "num_tokens": 781883371.0, "step": 20490 }, { "epoch": 2.6066658185981426, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.25963020324707, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8725166320800781, "num_tokens": 781919669.0, "step": 20491 }, { "epoch": 2.606793028876733, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.313836097717285, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8794706463813782, "num_tokens": 781953212.0, "step": 20492 }, { "epoch": 2.6069202391553237, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.202083587646484, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8773245811462402, "num_tokens": 781996824.0, "step": 20493 }, { "epoch": 2.607047449433914, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.21840763092041, "learning_rate": 1e-06, "loss": 0.335, "mean_token_accuracy": 0.8925034999847412, "num_tokens": 782035350.0, "step": 20494 }, { "epoch": 2.6071746597125047, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.314784049987793, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8683150410652161, "num_tokens": 782073217.0, "step": 20495 }, { "epoch": 2.6073018699910953, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.241028785705566, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.867082953453064, "num_tokens": 782110777.0, "step": 20496 }, { "epoch": 2.607429080269686, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.315028190612793, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8671870231628418, "num_tokens": 782153300.0, "step": 20497 }, { "epoch": 2.6075562905482763, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.251781463623047, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8786678314208984, "num_tokens": 782191945.0, "step": 20498 }, { "epoch": 2.607683500826867, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.267285346984863, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8841771483421326, "num_tokens": 782230472.0, "step": 20499 }, { "epoch": 2.6078107111054574, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.181888580322266, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8647595643997192, "num_tokens": 782263868.0, "step": 20500 }, { "epoch": 2.607937921384048, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.178668022155762, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8765038847923279, "num_tokens": 782299781.0, "step": 20501 }, { "epoch": 2.6080651316626384, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.250163078308105, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.866678774356842, "num_tokens": 782343376.0, "step": 20502 }, { "epoch": 2.608192341941229, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.181924819946289, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8855035305023193, "num_tokens": 782389894.0, "step": 20503 }, { "epoch": 2.6083195522198195, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.192737579345703, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8781594038009644, "num_tokens": 782430055.0, "step": 20504 }, { "epoch": 2.60844676249841, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.290703773498535, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8747318983078003, "num_tokens": 782467211.0, "step": 20505 }, { "epoch": 2.6085739727770005, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.18022346496582, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8649441599845886, "num_tokens": 782505913.0, "step": 20506 }, { "epoch": 2.608701183055591, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.280298233032227, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8730509281158447, "num_tokens": 782541858.0, "step": 20507 }, { "epoch": 2.608828393334181, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.218330383300781, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8734419345855713, "num_tokens": 782575854.0, "step": 20508 }, { "epoch": 2.608955603612772, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.249443054199219, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8756167888641357, "num_tokens": 782614005.0, "step": 20509 }, { "epoch": 2.609082813891362, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.232975006103516, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8810313940048218, "num_tokens": 782650147.0, "step": 20510 }, { "epoch": 2.609210024169953, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.223461151123047, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.872829020023346, "num_tokens": 782683215.0, "step": 20511 }, { "epoch": 2.6093372344485433, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.247018814086914, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8794379830360413, "num_tokens": 782721184.0, "step": 20512 }, { "epoch": 2.609464444727134, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.163209915161133, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8684791922569275, "num_tokens": 782759891.0, "step": 20513 }, { "epoch": 2.6095916550057243, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.193038940429688, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8635932207107544, "num_tokens": 782796398.0, "step": 20514 }, { "epoch": 2.609718865284315, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.22412395477295, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.883112907409668, "num_tokens": 782835968.0, "step": 20515 }, { "epoch": 2.6098460755629054, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.240060806274414, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.884620189666748, "num_tokens": 782871367.0, "step": 20516 }, { "epoch": 2.609973285841496, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.199080467224121, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.8937212228775024, "num_tokens": 782905766.0, "step": 20517 }, { "epoch": 2.6101004961200864, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.235090255737305, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8779959678649902, "num_tokens": 782942256.0, "step": 20518 }, { "epoch": 2.610227706398677, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.523331642150879, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8842412829399109, "num_tokens": 782976156.0, "step": 20519 }, { "epoch": 2.6103549166772675, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.204588890075684, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8822857141494751, "num_tokens": 783011450.0, "step": 20520 }, { "epoch": 2.610482126955858, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.127777099609375, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8787280321121216, "num_tokens": 783050355.0, "step": 20521 }, { "epoch": 2.6106093372344485, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.219759941101074, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.883926510810852, "num_tokens": 783091621.0, "step": 20522 }, { "epoch": 2.610736547513039, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.260414123535156, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8636296987533569, "num_tokens": 783130471.0, "step": 20523 }, { "epoch": 2.6108637577916296, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.186336517333984, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8788996934890747, "num_tokens": 783168340.0, "step": 20524 }, { "epoch": 2.61099096807022, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.273510932922363, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8557785749435425, "num_tokens": 783207279.0, "step": 20525 }, { "epoch": 2.6111181783488107, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.141331672668457, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8669637441635132, "num_tokens": 783243010.0, "step": 20526 }, { "epoch": 2.611245388627401, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.274637222290039, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.863805890083313, "num_tokens": 783287548.0, "step": 20527 }, { "epoch": 2.6113725989059917, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.238977432250977, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8711597919464111, "num_tokens": 783331549.0, "step": 20528 }, { "epoch": 2.6114998091845822, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.278005599975586, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.883270263671875, "num_tokens": 783369445.0, "step": 20529 }, { "epoch": 2.6116270194631728, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.232871055603027, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8691622614860535, "num_tokens": 783413218.0, "step": 20530 }, { "epoch": 2.6117542297417633, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.260658264160156, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8740290403366089, "num_tokens": 783449944.0, "step": 20531 }, { "epoch": 2.611881440020354, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.323860168457031, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8752317428588867, "num_tokens": 783486142.0, "step": 20532 }, { "epoch": 2.612008650298944, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.226011276245117, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8810631036758423, "num_tokens": 783521170.0, "step": 20533 }, { "epoch": 2.612135860577535, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.317401885986328, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8723406791687012, "num_tokens": 783563902.0, "step": 20534 }, { "epoch": 2.612263070856125, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.221343994140625, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8778191208839417, "num_tokens": 783597655.0, "step": 20535 }, { "epoch": 2.612390281134716, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.270628929138184, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8633931875228882, "num_tokens": 783635233.0, "step": 20536 }, { "epoch": 2.612517491413306, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.206598281860352, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8753922581672668, "num_tokens": 783679677.0, "step": 20537 }, { "epoch": 2.6126447016918966, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.198081016540527, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8802971839904785, "num_tokens": 783720608.0, "step": 20538 }, { "epoch": 2.612771911970487, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.31042766571045, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8840707540512085, "num_tokens": 783753198.0, "step": 20539 }, { "epoch": 2.6128991222490776, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.207554817199707, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8731521368026733, "num_tokens": 783791068.0, "step": 20540 }, { "epoch": 2.613026332527668, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.268362045288086, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8746548891067505, "num_tokens": 783831319.0, "step": 20541 }, { "epoch": 2.6131535428062587, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.195258140563965, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8775842189788818, "num_tokens": 783867537.0, "step": 20542 }, { "epoch": 2.613280753084849, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.238959312438965, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8690680265426636, "num_tokens": 783907787.0, "step": 20543 }, { "epoch": 2.6134079633634397, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.286171913146973, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8681830763816833, "num_tokens": 783948558.0, "step": 20544 }, { "epoch": 2.6135351736420303, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.248100280761719, "learning_rate": 1e-06, "loss": 0.347, "mean_token_accuracy": 0.8920082449913025, "num_tokens": 783986726.0, "step": 20545 }, { "epoch": 2.613662383920621, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.245772361755371, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8567931056022644, "num_tokens": 784023449.0, "step": 20546 }, { "epoch": 2.6137895941992113, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.238629341125488, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8906899690628052, "num_tokens": 784059529.0, "step": 20547 }, { "epoch": 2.613916804477802, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.293329238891602, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8750206232070923, "num_tokens": 784103530.0, "step": 20548 }, { "epoch": 2.6140440147563924, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.299028396606445, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8686911463737488, "num_tokens": 784141721.0, "step": 20549 }, { "epoch": 2.614171225034983, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.267492294311523, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8740038275718689, "num_tokens": 784175868.0, "step": 20550 }, { "epoch": 2.6142984353135734, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.297794342041016, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8851901292800903, "num_tokens": 784210646.0, "step": 20551 }, { "epoch": 2.614425645592164, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.250631332397461, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8800452947616577, "num_tokens": 784255588.0, "step": 20552 }, { "epoch": 2.6145528558707545, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.301891326904297, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8767483234405518, "num_tokens": 784292877.0, "step": 20553 }, { "epoch": 2.614680066149345, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2211332321167, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8781933188438416, "num_tokens": 784336150.0, "step": 20554 }, { "epoch": 2.6148072764279355, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.338568687438965, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8866807818412781, "num_tokens": 784371547.0, "step": 20555 }, { "epoch": 2.6149344867065256, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.266679763793945, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8762784004211426, "num_tokens": 784411473.0, "step": 20556 }, { "epoch": 2.6150616969851166, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.233282089233398, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8649547696113586, "num_tokens": 784456713.0, "step": 20557 }, { "epoch": 2.6151889072637067, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.224106788635254, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8885459899902344, "num_tokens": 784493685.0, "step": 20558 }, { "epoch": 2.6153161175422976, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.293251037597656, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8857107162475586, "num_tokens": 784534641.0, "step": 20559 }, { "epoch": 2.6154433278208877, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.230467796325684, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.8521444201469421, "num_tokens": 784576184.0, "step": 20560 }, { "epoch": 2.6155705380994787, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.206660270690918, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8794533014297485, "num_tokens": 784613851.0, "step": 20561 }, { "epoch": 2.615697748378069, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.176620483398438, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8767830729484558, "num_tokens": 784654133.0, "step": 20562 }, { "epoch": 2.6158249586566593, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.256000518798828, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8754780888557434, "num_tokens": 784690529.0, "step": 20563 }, { "epoch": 2.61595216893525, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.187882423400879, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.873986005783081, "num_tokens": 784731972.0, "step": 20564 }, { "epoch": 2.6160793792138404, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.220772743225098, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8706030249595642, "num_tokens": 784773275.0, "step": 20565 }, { "epoch": 2.616206589492431, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.278491020202637, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8601109981536865, "num_tokens": 784811468.0, "step": 20566 }, { "epoch": 2.6163337997710214, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.224620819091797, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.876915693283081, "num_tokens": 784850611.0, "step": 20567 }, { "epoch": 2.616461010049612, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.273597717285156, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8810406923294067, "num_tokens": 784886502.0, "step": 20568 }, { "epoch": 2.6165882203282025, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.178506851196289, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.866131603717804, "num_tokens": 784924489.0, "step": 20569 }, { "epoch": 2.616715430606793, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.268893241882324, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8800164461135864, "num_tokens": 784957224.0, "step": 20570 }, { "epoch": 2.6168426408853835, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.255687713623047, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8508365750312805, "num_tokens": 784997754.0, "step": 20571 }, { "epoch": 2.616969851163974, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.183084487915039, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8612782955169678, "num_tokens": 785033932.0, "step": 20572 }, { "epoch": 2.6170970614425646, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.358009338378906, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8714907169342041, "num_tokens": 785073190.0, "step": 20573 }, { "epoch": 2.617224271721155, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.234208106994629, "learning_rate": 1e-06, "loss": 0.344, "mean_token_accuracy": 0.890868067741394, "num_tokens": 785108397.0, "step": 20574 }, { "epoch": 2.6173514819997457, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.270713806152344, "learning_rate": 1e-06, "loss": 0.3234, "mean_token_accuracy": 0.9007563591003418, "num_tokens": 785142671.0, "step": 20575 }, { "epoch": 2.617478692278336, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.257481575012207, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8732692003250122, "num_tokens": 785183467.0, "step": 20576 }, { "epoch": 2.6176059025569267, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.278770446777344, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8638340830802917, "num_tokens": 785222072.0, "step": 20577 }, { "epoch": 2.6177331128355172, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.329771041870117, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8636441826820374, "num_tokens": 785260458.0, "step": 20578 }, { "epoch": 2.6178603231141078, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.252424240112305, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8609182238578796, "num_tokens": 785302434.0, "step": 20579 }, { "epoch": 2.6179875333926983, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.302984237670898, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8604875802993774, "num_tokens": 785338251.0, "step": 20580 }, { "epoch": 2.6181147436712884, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.23738956451416, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8700738549232483, "num_tokens": 785384789.0, "step": 20581 }, { "epoch": 2.6182419539498794, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.30583381652832, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8749257326126099, "num_tokens": 785424978.0, "step": 20582 }, { "epoch": 2.6183691642284694, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.256524085998535, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.866531252861023, "num_tokens": 785464421.0, "step": 20583 }, { "epoch": 2.6184963745070604, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.260087966918945, "learning_rate": 1e-06, "loss": 0.3335, "mean_token_accuracy": 0.8928398489952087, "num_tokens": 785496166.0, "step": 20584 }, { "epoch": 2.6186235847856505, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.370593070983887, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8597996234893799, "num_tokens": 785528504.0, "step": 20585 }, { "epoch": 2.6187507950642415, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.247884750366211, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.877147912979126, "num_tokens": 785570896.0, "step": 20586 }, { "epoch": 2.6188780053428315, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.270347595214844, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8702954053878784, "num_tokens": 785612464.0, "step": 20587 }, { "epoch": 2.619005215621422, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.277676582336426, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8742886185646057, "num_tokens": 785655500.0, "step": 20588 }, { "epoch": 2.6191324259000126, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.251710891723633, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8765882253646851, "num_tokens": 785691186.0, "step": 20589 }, { "epoch": 2.619259636178603, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.338088035583496, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8865147829055786, "num_tokens": 785728575.0, "step": 20590 }, { "epoch": 2.6193868464571937, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.243863105773926, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8716030716896057, "num_tokens": 785765782.0, "step": 20591 }, { "epoch": 2.619514056735784, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.312114715576172, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8790377974510193, "num_tokens": 785800130.0, "step": 20592 }, { "epoch": 2.6196412670143747, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.192092895507812, "learning_rate": 1e-06, "loss": 0.3126, "mean_token_accuracy": 0.9015350937843323, "num_tokens": 785836647.0, "step": 20593 }, { "epoch": 2.6197684772929652, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.309555053710938, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8690071105957031, "num_tokens": 785873897.0, "step": 20594 }, { "epoch": 2.6198956875715558, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.254176139831543, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8702107667922974, "num_tokens": 785914296.0, "step": 20595 }, { "epoch": 2.6200228978501463, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.287385940551758, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8624035120010376, "num_tokens": 785951426.0, "step": 20596 }, { "epoch": 2.620150108128737, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.237652778625488, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8764883279800415, "num_tokens": 785989906.0, "step": 20597 }, { "epoch": 2.6202773184073274, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.3096284866333, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8693060278892517, "num_tokens": 786024763.0, "step": 20598 }, { "epoch": 2.620404528685918, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.294618606567383, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8700669407844543, "num_tokens": 786063055.0, "step": 20599 }, { "epoch": 2.6205317389645084, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.226495742797852, "learning_rate": 1e-06, "loss": 0.4951, "mean_token_accuracy": 0.8484252691268921, "num_tokens": 786097675.0, "step": 20600 }, { "epoch": 2.620658949243099, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.348296165466309, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8856046199798584, "num_tokens": 786133721.0, "step": 20601 }, { "epoch": 2.6207861595216895, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.25367259979248, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8560447692871094, "num_tokens": 786179615.0, "step": 20602 }, { "epoch": 2.62091336980028, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.26405143737793, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8874166011810303, "num_tokens": 786214945.0, "step": 20603 }, { "epoch": 2.6210405800788705, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2345609664917, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8758875131607056, "num_tokens": 786250880.0, "step": 20604 }, { "epoch": 2.621167790357461, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.254021644592285, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8741194605827332, "num_tokens": 786286095.0, "step": 20605 }, { "epoch": 2.621295000636051, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.26626205444336, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8789026737213135, "num_tokens": 786319547.0, "step": 20606 }, { "epoch": 2.621422210914642, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.23177433013916, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8743350505828857, "num_tokens": 786358699.0, "step": 20607 }, { "epoch": 2.621549421193232, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.263651847839355, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8780415058135986, "num_tokens": 786399732.0, "step": 20608 }, { "epoch": 2.621676631471823, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.291354179382324, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8652524948120117, "num_tokens": 786437375.0, "step": 20609 }, { "epoch": 2.6218038417504133, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.279366493225098, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8639837503433228, "num_tokens": 786475538.0, "step": 20610 }, { "epoch": 2.621931052029004, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.27387809753418, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.876042366027832, "num_tokens": 786516802.0, "step": 20611 }, { "epoch": 2.6220582623075943, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.23324966430664, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8742495775222778, "num_tokens": 786558470.0, "step": 20612 }, { "epoch": 2.622185472586185, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.32154655456543, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8689085245132446, "num_tokens": 786601163.0, "step": 20613 }, { "epoch": 2.6223126828647754, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.235058784484863, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8813546895980835, "num_tokens": 786636741.0, "step": 20614 }, { "epoch": 2.622439893143366, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.313985824584961, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8629885315895081, "num_tokens": 786672931.0, "step": 20615 }, { "epoch": 2.6225671034219564, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.246452331542969, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8863945007324219, "num_tokens": 786709310.0, "step": 20616 }, { "epoch": 2.622694313700547, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.342586517333984, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8799488544464111, "num_tokens": 786747113.0, "step": 20617 }, { "epoch": 2.6228215239791375, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.36107063293457, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8730393648147583, "num_tokens": 786782033.0, "step": 20618 }, { "epoch": 2.622948734257728, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.2356538772583, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8749597072601318, "num_tokens": 786827229.0, "step": 20619 }, { "epoch": 2.6230759445363185, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.352108001708984, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8810760378837585, "num_tokens": 786861722.0, "step": 20620 }, { "epoch": 2.623203154814909, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.31801700592041, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8782554268836975, "num_tokens": 786896202.0, "step": 20621 }, { "epoch": 2.6233303650934996, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.358824729919434, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8761087656021118, "num_tokens": 786932154.0, "step": 20622 }, { "epoch": 2.62345757537209, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.283720016479492, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8658640384674072, "num_tokens": 786972184.0, "step": 20623 }, { "epoch": 2.6235847856506807, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.287508964538574, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8681045174598694, "num_tokens": 787015514.0, "step": 20624 }, { "epoch": 2.623711995929271, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.283136367797852, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8748621940612793, "num_tokens": 787054023.0, "step": 20625 }, { "epoch": 2.6238392062078617, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.329915046691895, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8711783289909363, "num_tokens": 787088858.0, "step": 20626 }, { "epoch": 2.6239664164864522, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.260035514831543, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8767732381820679, "num_tokens": 787124805.0, "step": 20627 }, { "epoch": 2.6240936267650428, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.27970027923584, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8862155079841614, "num_tokens": 787164922.0, "step": 20628 }, { "epoch": 2.6242208370436333, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.321290016174316, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8866913318634033, "num_tokens": 787204936.0, "step": 20629 }, { "epoch": 2.624348047322224, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.319055557250977, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8856192827224731, "num_tokens": 787238292.0, "step": 20630 }, { "epoch": 2.624475257600814, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.370420455932617, "learning_rate": 1e-06, "loss": 0.3364, "mean_token_accuracy": 0.8930215239524841, "num_tokens": 787269054.0, "step": 20631 }, { "epoch": 2.624602467879405, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.329961776733398, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8734704852104187, "num_tokens": 787310022.0, "step": 20632 }, { "epoch": 2.624729678157995, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.322538375854492, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8643730878829956, "num_tokens": 787350917.0, "step": 20633 }, { "epoch": 2.624856888436586, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.314858436584473, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.88318932056427, "num_tokens": 787388187.0, "step": 20634 }, { "epoch": 2.624984098715176, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.335311889648438, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8809427618980408, "num_tokens": 787424998.0, "step": 20635 }, { "epoch": 2.6251113089937665, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.255355834960938, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8860714435577393, "num_tokens": 787465456.0, "step": 20636 }, { "epoch": 2.625238519272357, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.303047180175781, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.872963547706604, "num_tokens": 787504797.0, "step": 20637 }, { "epoch": 2.6253657295509476, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.301253318786621, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8849154710769653, "num_tokens": 787542911.0, "step": 20638 }, { "epoch": 2.625492939829538, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.35044002532959, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8659337759017944, "num_tokens": 787576143.0, "step": 20639 }, { "epoch": 2.6256201501081287, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.330415725708008, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8769664764404297, "num_tokens": 787619912.0, "step": 20640 }, { "epoch": 2.625747360386719, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.256536483764648, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8644886016845703, "num_tokens": 787658181.0, "step": 20641 }, { "epoch": 2.6258745706653097, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.33597183227539, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8657907843589783, "num_tokens": 787694889.0, "step": 20642 }, { "epoch": 2.6260017809439002, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.298290252685547, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8871633410453796, "num_tokens": 787737633.0, "step": 20643 }, { "epoch": 2.6261289912224908, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.297529220581055, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8628673553466797, "num_tokens": 787779069.0, "step": 20644 }, { "epoch": 2.6262562015010813, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.356663703918457, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8620362877845764, "num_tokens": 787815544.0, "step": 20645 }, { "epoch": 2.626383411779672, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.346125602722168, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.869093656539917, "num_tokens": 787852684.0, "step": 20646 }, { "epoch": 2.6265106220582624, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.281235694885254, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8708974719047546, "num_tokens": 787893452.0, "step": 20647 }, { "epoch": 2.626637832336853, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.292082786560059, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8741134405136108, "num_tokens": 787929430.0, "step": 20648 }, { "epoch": 2.6267650426154434, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.284618377685547, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8719733953475952, "num_tokens": 787969768.0, "step": 20649 }, { "epoch": 2.626892252894034, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.269434928894043, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8737443089485168, "num_tokens": 788009605.0, "step": 20650 }, { "epoch": 2.6270194631726245, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.34453010559082, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8860326409339905, "num_tokens": 788044652.0, "step": 20651 }, { "epoch": 2.627146673451215, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.355719566345215, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8489540815353394, "num_tokens": 788085981.0, "step": 20652 }, { "epoch": 2.6272738837298055, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.26469898223877, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8712748885154724, "num_tokens": 788132196.0, "step": 20653 }, { "epoch": 2.6274010940083956, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.370689392089844, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8866324424743652, "num_tokens": 788160099.0, "step": 20654 }, { "epoch": 2.6275283042869866, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.283528327941895, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8729245662689209, "num_tokens": 788197264.0, "step": 20655 }, { "epoch": 2.6276555145655767, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.288985252380371, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8856698274612427, "num_tokens": 788232900.0, "step": 20656 }, { "epoch": 2.6277827248441676, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.250317573547363, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8683822154998779, "num_tokens": 788273720.0, "step": 20657 }, { "epoch": 2.6279099351227577, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.34087085723877, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8767441511154175, "num_tokens": 788308425.0, "step": 20658 }, { "epoch": 2.6280371454013487, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.298781394958496, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8662574291229248, "num_tokens": 788343701.0, "step": 20659 }, { "epoch": 2.628164355679939, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.392329216003418, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8588688373565674, "num_tokens": 788372789.0, "step": 20660 }, { "epoch": 2.6282915659585293, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.327119827270508, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8788954019546509, "num_tokens": 788408422.0, "step": 20661 }, { "epoch": 2.62841877623712, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.322226524353027, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.878210186958313, "num_tokens": 788446113.0, "step": 20662 }, { "epoch": 2.6285459865157104, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.322346687316895, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8686615824699402, "num_tokens": 788485009.0, "step": 20663 }, { "epoch": 2.628673196794301, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.306987762451172, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8775880932807922, "num_tokens": 788523446.0, "step": 20664 }, { "epoch": 2.6288004070728914, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.375894546508789, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8827790021896362, "num_tokens": 788557423.0, "step": 20665 }, { "epoch": 2.628927617351482, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.376803398132324, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8721022605895996, "num_tokens": 788588177.0, "step": 20666 }, { "epoch": 2.6290548276300725, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.266263008117676, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8755382895469666, "num_tokens": 788624877.0, "step": 20667 }, { "epoch": 2.629182037908663, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.376697540283203, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8846524357795715, "num_tokens": 788662532.0, "step": 20668 }, { "epoch": 2.6293092481872535, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.18920612335205, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8749643564224243, "num_tokens": 788702734.0, "step": 20669 }, { "epoch": 2.629436458465844, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.292357444763184, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8747350573539734, "num_tokens": 788738041.0, "step": 20670 }, { "epoch": 2.6295636687444346, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.317387580871582, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.875724196434021, "num_tokens": 788772998.0, "step": 20671 }, { "epoch": 2.629690879023025, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.273341178894043, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8593794107437134, "num_tokens": 788814310.0, "step": 20672 }, { "epoch": 2.6298180893016156, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.3073148727417, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8786769509315491, "num_tokens": 788852811.0, "step": 20673 }, { "epoch": 2.629945299580206, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.2307767868042, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.882143497467041, "num_tokens": 788888756.0, "step": 20674 }, { "epoch": 2.6300725098587967, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.313920974731445, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8777410387992859, "num_tokens": 788926292.0, "step": 20675 }, { "epoch": 2.6301997201373872, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.294466018676758, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8843992352485657, "num_tokens": 788965625.0, "step": 20676 }, { "epoch": 2.6303269304159778, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.325018882751465, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8787945508956909, "num_tokens": 789007390.0, "step": 20677 }, { "epoch": 2.6304541406945683, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.308612823486328, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8776429891586304, "num_tokens": 789047940.0, "step": 20678 }, { "epoch": 2.6305813509731584, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.315727233886719, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8745110630989075, "num_tokens": 789086321.0, "step": 20679 }, { "epoch": 2.6307085612517493, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.424695014953613, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8859139680862427, "num_tokens": 789121757.0, "step": 20680 }, { "epoch": 2.6308357715303394, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.317937850952148, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8882521390914917, "num_tokens": 789160076.0, "step": 20681 }, { "epoch": 2.6309629818089304, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.337817192077637, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8653479218482971, "num_tokens": 789197080.0, "step": 20682 }, { "epoch": 2.6310901920875205, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.257023811340332, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8690845966339111, "num_tokens": 789249876.0, "step": 20683 }, { "epoch": 2.6312174023661115, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.244120597839355, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8725154995918274, "num_tokens": 789290146.0, "step": 20684 }, { "epoch": 2.6313446126447015, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.321110725402832, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8781356811523438, "num_tokens": 789335610.0, "step": 20685 }, { "epoch": 2.631471822923292, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.335836410522461, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8702818751335144, "num_tokens": 789371956.0, "step": 20686 }, { "epoch": 2.6315990332018826, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.313776016235352, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8704375624656677, "num_tokens": 789415257.0, "step": 20687 }, { "epoch": 2.631726243480473, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.340240478515625, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8752596378326416, "num_tokens": 789452533.0, "step": 20688 }, { "epoch": 2.6318534537590637, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.410327911376953, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8713951110839844, "num_tokens": 789488780.0, "step": 20689 }, { "epoch": 2.631980664037654, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.311509132385254, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8717894554138184, "num_tokens": 789528169.0, "step": 20690 }, { "epoch": 2.6321078743162447, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.34351921081543, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.867486834526062, "num_tokens": 789568382.0, "step": 20691 }, { "epoch": 2.6322350845948352, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.274142265319824, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8721190690994263, "num_tokens": 789606895.0, "step": 20692 }, { "epoch": 2.6323622948734258, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.382525444030762, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8597484230995178, "num_tokens": 789649106.0, "step": 20693 }, { "epoch": 2.6324895051520163, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.379764556884766, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.878332257270813, "num_tokens": 789689449.0, "step": 20694 }, { "epoch": 2.632616715430607, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.271232604980469, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8854402303695679, "num_tokens": 789726179.0, "step": 20695 }, { "epoch": 2.6327439257091974, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.26360034942627, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8576506972312927, "num_tokens": 789764858.0, "step": 20696 }, { "epoch": 2.632871135987788, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.325634002685547, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8623499274253845, "num_tokens": 789799196.0, "step": 20697 }, { "epoch": 2.6329983462663784, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.280719757080078, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8719367980957031, "num_tokens": 789840498.0, "step": 20698 }, { "epoch": 2.633125556544969, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.3240966796875, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8774400949478149, "num_tokens": 789874269.0, "step": 20699 }, { "epoch": 2.6332527668235595, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.281271934509277, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8754801154136658, "num_tokens": 789914938.0, "step": 20700 }, { "epoch": 2.63337997710215, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.293316841125488, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8744773864746094, "num_tokens": 789959219.0, "step": 20701 }, { "epoch": 2.6335071873807405, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.281698226928711, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8678140044212341, "num_tokens": 789993707.0, "step": 20702 }, { "epoch": 2.633634397659331, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.242809295654297, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8549479842185974, "num_tokens": 790034285.0, "step": 20703 }, { "epoch": 2.633761607937921, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.344225883483887, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8765569925308228, "num_tokens": 790069714.0, "step": 20704 }, { "epoch": 2.633888818216512, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.315755844116211, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8728984594345093, "num_tokens": 790110733.0, "step": 20705 }, { "epoch": 2.634016028495102, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.329113006591797, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8789709210395813, "num_tokens": 790144468.0, "step": 20706 }, { "epoch": 2.634143238773693, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.27808666229248, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8712369203567505, "num_tokens": 790186881.0, "step": 20707 }, { "epoch": 2.6342704490522832, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.370684623718262, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8810422420501709, "num_tokens": 790225083.0, "step": 20708 }, { "epoch": 2.6343976593308738, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.226218223571777, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8739116787910461, "num_tokens": 790266822.0, "step": 20709 }, { "epoch": 2.6345248696094643, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.23194694519043, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8705663681030273, "num_tokens": 790305948.0, "step": 20710 }, { "epoch": 2.634652079888055, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.357799530029297, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8763223886489868, "num_tokens": 790341639.0, "step": 20711 }, { "epoch": 2.6347792901666454, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.24482250213623, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8832845687866211, "num_tokens": 790377442.0, "step": 20712 }, { "epoch": 2.634906500445236, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.289830207824707, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8740711808204651, "num_tokens": 790414911.0, "step": 20713 }, { "epoch": 2.6350337107238264, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.339192390441895, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8719943761825562, "num_tokens": 790453968.0, "step": 20714 }, { "epoch": 2.635160921002417, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.282495498657227, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8701688647270203, "num_tokens": 790493216.0, "step": 20715 }, { "epoch": 2.6352881312810075, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.344954490661621, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8699690103530884, "num_tokens": 790529380.0, "step": 20716 }, { "epoch": 2.635415341559598, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.235406875610352, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8689262866973877, "num_tokens": 790569568.0, "step": 20717 }, { "epoch": 2.6355425518381885, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.343334197998047, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8808227777481079, "num_tokens": 790602269.0, "step": 20718 }, { "epoch": 2.635669762116779, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.284819602966309, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.86338210105896, "num_tokens": 790639723.0, "step": 20719 }, { "epoch": 2.6357969723953696, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.32628345489502, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8605189323425293, "num_tokens": 790676103.0, "step": 20720 }, { "epoch": 2.63592418267396, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.236615180969238, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8759034872055054, "num_tokens": 790712118.0, "step": 20721 }, { "epoch": 2.6360513929525506, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.294763565063477, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8736964464187622, "num_tokens": 790747079.0, "step": 20722 }, { "epoch": 2.636178603231141, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.31988525390625, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8642003536224365, "num_tokens": 790781741.0, "step": 20723 }, { "epoch": 2.6363058135097317, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.3441801071167, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8730643391609192, "num_tokens": 790820243.0, "step": 20724 }, { "epoch": 2.6364330237883222, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.347169876098633, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8885963559150696, "num_tokens": 790851492.0, "step": 20725 }, { "epoch": 2.6365602340669128, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.324405670166016, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.879540205001831, "num_tokens": 790882293.0, "step": 20726 }, { "epoch": 2.636687444345503, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2921724319458, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8703720569610596, "num_tokens": 790922188.0, "step": 20727 }, { "epoch": 2.636814654624094, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.366044998168945, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.884190022945404, "num_tokens": 790960675.0, "step": 20728 }, { "epoch": 2.636941864902684, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.385985374450684, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8829975128173828, "num_tokens": 790999352.0, "step": 20729 }, { "epoch": 2.637069075181275, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.296207427978516, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8746501803398132, "num_tokens": 791035898.0, "step": 20730 }, { "epoch": 2.637196285459865, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.331686019897461, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8500080108642578, "num_tokens": 791079731.0, "step": 20731 }, { "epoch": 2.637323495738456, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.248343467712402, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8867402076721191, "num_tokens": 791118565.0, "step": 20732 }, { "epoch": 2.637450706017046, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.323202133178711, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8753043413162231, "num_tokens": 791154876.0, "step": 20733 }, { "epoch": 2.6375779162956365, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.327058792114258, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8741124868392944, "num_tokens": 791196003.0, "step": 20734 }, { "epoch": 2.637705126574227, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.241888046264648, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.872234582901001, "num_tokens": 791236462.0, "step": 20735 }, { "epoch": 2.6378323368528176, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.362425804138184, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8610894680023193, "num_tokens": 791271918.0, "step": 20736 }, { "epoch": 2.637959547131408, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.225988388061523, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8853614926338196, "num_tokens": 791309813.0, "step": 20737 }, { "epoch": 2.6380867574099987, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.342576026916504, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8829295039176941, "num_tokens": 791338854.0, "step": 20738 }, { "epoch": 2.638213967688589, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.31804370880127, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.879485011100769, "num_tokens": 791376580.0, "step": 20739 }, { "epoch": 2.6383411779671797, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.277216911315918, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8697588443756104, "num_tokens": 791415931.0, "step": 20740 }, { "epoch": 2.6384683882457702, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.296204566955566, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8702297210693359, "num_tokens": 791448201.0, "step": 20741 }, { "epoch": 2.6385955985243608, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.295624732971191, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8672159314155579, "num_tokens": 791489183.0, "step": 20742 }, { "epoch": 2.6387228088029513, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.346723556518555, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8748129606246948, "num_tokens": 791528414.0, "step": 20743 }, { "epoch": 2.638850019081542, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.279574394226074, "learning_rate": 1e-06, "loss": 0.34, "mean_token_accuracy": 0.8935871720314026, "num_tokens": 791566379.0, "step": 20744 }, { "epoch": 2.6389772293601323, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.372824668884277, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8773573637008667, "num_tokens": 791599789.0, "step": 20745 }, { "epoch": 2.639104439638723, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.363330841064453, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8610610365867615, "num_tokens": 791642333.0, "step": 20746 }, { "epoch": 2.6392316499173134, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.259763717651367, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8751455545425415, "num_tokens": 791687190.0, "step": 20747 }, { "epoch": 2.639358860195904, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.332560539245605, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8594991564750671, "num_tokens": 791722806.0, "step": 20748 }, { "epoch": 2.6394860704744945, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.325715065002441, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8592010736465454, "num_tokens": 791757726.0, "step": 20749 }, { "epoch": 2.639613280753085, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.361204147338867, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8817003965377808, "num_tokens": 791795160.0, "step": 20750 }, { "epoch": 2.6397404910316755, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.350053787231445, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8851042985916138, "num_tokens": 791831682.0, "step": 20751 }, { "epoch": 2.6398677013102656, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.29342269897461, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8831733465194702, "num_tokens": 791863230.0, "step": 20752 }, { "epoch": 2.6399949115888566, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.308745384216309, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8867191672325134, "num_tokens": 791904323.0, "step": 20753 }, { "epoch": 2.6401221218674467, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.277464866638184, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.866949200630188, "num_tokens": 791951836.0, "step": 20754 }, { "epoch": 2.6402493321460376, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.349177360534668, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8755444288253784, "num_tokens": 791989782.0, "step": 20755 }, { "epoch": 2.6403765424246277, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.329703330993652, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.874954879283905, "num_tokens": 792024038.0, "step": 20756 }, { "epoch": 2.6405037527032187, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.291889190673828, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.874732494354248, "num_tokens": 792067893.0, "step": 20757 }, { "epoch": 2.6406309629818088, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.297683715820312, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8830268383026123, "num_tokens": 792108593.0, "step": 20758 }, { "epoch": 2.6407581732603993, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.318572044372559, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8862780928611755, "num_tokens": 792141201.0, "step": 20759 }, { "epoch": 2.64088538353899, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.277759552001953, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.894091010093689, "num_tokens": 792182908.0, "step": 20760 }, { "epoch": 2.6410125938175804, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.325495719909668, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8752185106277466, "num_tokens": 792225223.0, "step": 20761 }, { "epoch": 2.641139804096171, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.358378410339355, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8574748039245605, "num_tokens": 792258131.0, "step": 20762 }, { "epoch": 2.6412670143747614, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.304931640625, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8802223205566406, "num_tokens": 792291429.0, "step": 20763 }, { "epoch": 2.641394224653352, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.280143737792969, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8703368306159973, "num_tokens": 792327777.0, "step": 20764 }, { "epoch": 2.6415214349319425, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.340641975402832, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8767828941345215, "num_tokens": 792367918.0, "step": 20765 }, { "epoch": 2.641648645210533, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.33507251739502, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8778622150421143, "num_tokens": 792405550.0, "step": 20766 }, { "epoch": 2.6417758554891235, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.315413475036621, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8766661286354065, "num_tokens": 792440948.0, "step": 20767 }, { "epoch": 2.641903065767714, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.306106567382812, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8818639516830444, "num_tokens": 792488065.0, "step": 20768 }, { "epoch": 2.6420302760463046, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.322647094726562, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8817501068115234, "num_tokens": 792522845.0, "step": 20769 }, { "epoch": 2.642157486324895, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.312592506408691, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8848453760147095, "num_tokens": 792561118.0, "step": 20770 }, { "epoch": 2.6422846966034856, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.28223705291748, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8745135068893433, "num_tokens": 792600125.0, "step": 20771 }, { "epoch": 2.642411906882076, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.309142112731934, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8808659315109253, "num_tokens": 792633856.0, "step": 20772 }, { "epoch": 2.6425391171606667, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.363726615905762, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8714240789413452, "num_tokens": 792672427.0, "step": 20773 }, { "epoch": 2.6426663274392572, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.291082382202148, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.853054404258728, "num_tokens": 792713856.0, "step": 20774 }, { "epoch": 2.6427935377178478, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.404960632324219, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8674954771995544, "num_tokens": 792757602.0, "step": 20775 }, { "epoch": 2.6429207479964383, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.309566497802734, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8664195537567139, "num_tokens": 792797067.0, "step": 20776 }, { "epoch": 2.6430479582750284, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.357025146484375, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8784083127975464, "num_tokens": 792835748.0, "step": 20777 }, { "epoch": 2.6431751685536193, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.298169136047363, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8758598566055298, "num_tokens": 792875039.0, "step": 20778 }, { "epoch": 2.6433023788322094, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.335424423217773, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8800030946731567, "num_tokens": 792916580.0, "step": 20779 }, { "epoch": 2.6434295891108004, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.298436164855957, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.876443088054657, "num_tokens": 792952507.0, "step": 20780 }, { "epoch": 2.6435567993893905, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.307332992553711, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8781644105911255, "num_tokens": 792994046.0, "step": 20781 }, { "epoch": 2.6436840096679814, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.31069564819336, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.8926150798797607, "num_tokens": 793033855.0, "step": 20782 }, { "epoch": 2.6438112199465715, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.331391334533691, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8785967826843262, "num_tokens": 793072226.0, "step": 20783 }, { "epoch": 2.643938430225162, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.277191162109375, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8786575794219971, "num_tokens": 793110884.0, "step": 20784 }, { "epoch": 2.6440656405037526, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.32691478729248, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8567283153533936, "num_tokens": 793151885.0, "step": 20785 }, { "epoch": 2.644192850782343, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.281861305236816, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8528931736946106, "num_tokens": 793187253.0, "step": 20786 }, { "epoch": 2.6443200610609336, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.261956214904785, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8799687623977661, "num_tokens": 793231249.0, "step": 20787 }, { "epoch": 2.644447271339524, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.284459114074707, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8857875466346741, "num_tokens": 793269918.0, "step": 20788 }, { "epoch": 2.6445744816181147, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.2754545211792, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8854349851608276, "num_tokens": 793306856.0, "step": 20789 }, { "epoch": 2.6447016918967052, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.353699684143066, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.865618109703064, "num_tokens": 793347866.0, "step": 20790 }, { "epoch": 2.6448289021752958, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.19134521484375, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8732858896255493, "num_tokens": 793391037.0, "step": 20791 }, { "epoch": 2.6449561124538863, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.372065544128418, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8616310358047485, "num_tokens": 793423560.0, "step": 20792 }, { "epoch": 2.645083322732477, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.344893455505371, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8817760944366455, "num_tokens": 793460215.0, "step": 20793 }, { "epoch": 2.6452105330110673, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.393362045288086, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8631525039672852, "num_tokens": 793492671.0, "step": 20794 }, { "epoch": 2.645337743289658, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.335443496704102, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.873467206954956, "num_tokens": 793529526.0, "step": 20795 }, { "epoch": 2.6454649535682484, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.311362266540527, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8835180401802063, "num_tokens": 793566748.0, "step": 20796 }, { "epoch": 2.645592163846839, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.336891174316406, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.889331042766571, "num_tokens": 793600703.0, "step": 20797 }, { "epoch": 2.6457193741254295, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.420431137084961, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8739669919013977, "num_tokens": 793639226.0, "step": 20798 }, { "epoch": 2.64584658440402, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.26395320892334, "learning_rate": 1e-06, "loss": 0.3494, "mean_token_accuracy": 0.88796067237854, "num_tokens": 793676147.0, "step": 20799 }, { "epoch": 2.6459737946826105, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.338748931884766, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.885155975818634, "num_tokens": 793715778.0, "step": 20800 }, { "epoch": 2.646101004961201, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.305408477783203, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8899269104003906, "num_tokens": 793752096.0, "step": 20801 }, { "epoch": 2.646228215239791, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.315606117248535, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8694547414779663, "num_tokens": 793789809.0, "step": 20802 }, { "epoch": 2.646355425518382, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.339348793029785, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8711448907852173, "num_tokens": 793826505.0, "step": 20803 }, { "epoch": 2.646482635796972, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.354403495788574, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8746927976608276, "num_tokens": 793868360.0, "step": 20804 }, { "epoch": 2.646609846075563, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.263467788696289, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8765576481819153, "num_tokens": 793906566.0, "step": 20805 }, { "epoch": 2.6467370563541532, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.360511779785156, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.872528612613678, "num_tokens": 793947814.0, "step": 20806 }, { "epoch": 2.6468642666327438, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.282538414001465, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8804798722267151, "num_tokens": 793987391.0, "step": 20807 }, { "epoch": 2.6469914769113343, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.36412525177002, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8858449459075928, "num_tokens": 794024045.0, "step": 20808 }, { "epoch": 2.647118687189925, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.259811401367188, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8793550729751587, "num_tokens": 794063405.0, "step": 20809 }, { "epoch": 2.6472458974685154, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.384102821350098, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8519205451011658, "num_tokens": 794099729.0, "step": 20810 }, { "epoch": 2.647373107747106, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.279176712036133, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8672067523002625, "num_tokens": 794139669.0, "step": 20811 }, { "epoch": 2.6475003180256964, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.308808326721191, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8763341903686523, "num_tokens": 794176906.0, "step": 20812 }, { "epoch": 2.647627528304287, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.303244590759277, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8778249621391296, "num_tokens": 794216594.0, "step": 20813 }, { "epoch": 2.6477547385828775, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.252176284790039, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8779538869857788, "num_tokens": 794248379.0, "step": 20814 }, { "epoch": 2.647881948861468, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.332941055297852, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8528724908828735, "num_tokens": 794293604.0, "step": 20815 }, { "epoch": 2.6480091591400585, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.313054084777832, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8652622699737549, "num_tokens": 794332084.0, "step": 20816 }, { "epoch": 2.648136369418649, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.214437484741211, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8901717662811279, "num_tokens": 794369158.0, "step": 20817 }, { "epoch": 2.6482635796972396, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.360685348510742, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8781448602676392, "num_tokens": 794410531.0, "step": 20818 }, { "epoch": 2.64839078997583, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.32343578338623, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8760988712310791, "num_tokens": 794457556.0, "step": 20819 }, { "epoch": 2.6485180002544206, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.322859764099121, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.863684892654419, "num_tokens": 794492737.0, "step": 20820 }, { "epoch": 2.648645210533011, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.329133033752441, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8851792812347412, "num_tokens": 794532589.0, "step": 20821 }, { "epoch": 2.6487724208116017, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.33432674407959, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.864651620388031, "num_tokens": 794566982.0, "step": 20822 }, { "epoch": 2.648899631090192, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.308609008789062, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8590754866600037, "num_tokens": 794604278.0, "step": 20823 }, { "epoch": 2.6490268413687827, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.362166404724121, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8839137554168701, "num_tokens": 794643785.0, "step": 20824 }, { "epoch": 2.649154051647373, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.215899467468262, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8840551972389221, "num_tokens": 794678302.0, "step": 20825 }, { "epoch": 2.649281261925964, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.36212158203125, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8653782606124878, "num_tokens": 794715732.0, "step": 20826 }, { "epoch": 2.649408472204554, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.294342994689941, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8785132169723511, "num_tokens": 794754359.0, "step": 20827 }, { "epoch": 2.649535682483145, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.300795555114746, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8713721036911011, "num_tokens": 794795770.0, "step": 20828 }, { "epoch": 2.649662892761735, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.28249740600586, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8677371144294739, "num_tokens": 794835654.0, "step": 20829 }, { "epoch": 2.649790103040326, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.387748718261719, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8768973350524902, "num_tokens": 794872527.0, "step": 20830 }, { "epoch": 2.649917313318916, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.305042266845703, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.879203200340271, "num_tokens": 794918577.0, "step": 20831 }, { "epoch": 2.6500445235975065, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.354101181030273, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8847995400428772, "num_tokens": 794953566.0, "step": 20832 }, { "epoch": 2.650171733876097, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.37397289276123, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8815425634384155, "num_tokens": 794990547.0, "step": 20833 }, { "epoch": 2.6502989441546876, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.297075271606445, "learning_rate": 1e-06, "loss": 0.3447, "mean_token_accuracy": 0.8922736048698425, "num_tokens": 795028419.0, "step": 20834 }, { "epoch": 2.650426154433278, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.393256187438965, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8904372453689575, "num_tokens": 795064303.0, "step": 20835 }, { "epoch": 2.6505533647118686, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.235921859741211, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8543872833251953, "num_tokens": 795104280.0, "step": 20836 }, { "epoch": 2.650680574990459, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.38936996459961, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8667216300964355, "num_tokens": 795142547.0, "step": 20837 }, { "epoch": 2.6508077852690497, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.26177978515625, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8728224635124207, "num_tokens": 795186193.0, "step": 20838 }, { "epoch": 2.6509349955476402, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.37143325805664, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.8925475478172302, "num_tokens": 795215453.0, "step": 20839 }, { "epoch": 2.6510622058262308, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.34679126739502, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.870145320892334, "num_tokens": 795252820.0, "step": 20840 }, { "epoch": 2.6511894161048213, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.350563049316406, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8730737566947937, "num_tokens": 795290000.0, "step": 20841 }, { "epoch": 2.651316626383412, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.300374031066895, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.8925937414169312, "num_tokens": 795330389.0, "step": 20842 }, { "epoch": 2.6514438366620023, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.357946395874023, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8725813627243042, "num_tokens": 795366693.0, "step": 20843 }, { "epoch": 2.651571046940593, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.41146469116211, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8765112161636353, "num_tokens": 795405748.0, "step": 20844 }, { "epoch": 2.6516982572191834, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.268339157104492, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8850409388542175, "num_tokens": 795446179.0, "step": 20845 }, { "epoch": 2.651825467497774, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.369019508361816, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8700789213180542, "num_tokens": 795488234.0, "step": 20846 }, { "epoch": 2.6519526777763645, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.293252944946289, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.872395396232605, "num_tokens": 795526508.0, "step": 20847 }, { "epoch": 2.652079888054955, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.304615020751953, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.881486713886261, "num_tokens": 795566299.0, "step": 20848 }, { "epoch": 2.6522070983335455, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.316961288452148, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8710147738456726, "num_tokens": 795603892.0, "step": 20849 }, { "epoch": 2.6523343086121356, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.393404006958008, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8825110793113708, "num_tokens": 795639622.0, "step": 20850 }, { "epoch": 2.6524615188907266, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.251912117004395, "learning_rate": 1e-06, "loss": 0.3293, "mean_token_accuracy": 0.89589923620224, "num_tokens": 795677299.0, "step": 20851 }, { "epoch": 2.6525887291693167, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.374856948852539, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8756381273269653, "num_tokens": 795714304.0, "step": 20852 }, { "epoch": 2.6527159394479076, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.266876220703125, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8712172508239746, "num_tokens": 795752330.0, "step": 20853 }, { "epoch": 2.6528431497264977, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.34032917022705, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8552275896072388, "num_tokens": 795789503.0, "step": 20854 }, { "epoch": 2.6529703600050887, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.217681884765625, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8741186857223511, "num_tokens": 795828449.0, "step": 20855 }, { "epoch": 2.6530975702836788, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.36741828918457, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8755784034729004, "num_tokens": 795868454.0, "step": 20856 }, { "epoch": 2.6532247805622693, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.353241920471191, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8620020151138306, "num_tokens": 795904629.0, "step": 20857 }, { "epoch": 2.65335199084086, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.386570930480957, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8614920377731323, "num_tokens": 795943197.0, "step": 20858 }, { "epoch": 2.6534792011194503, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.309781074523926, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8713164925575256, "num_tokens": 795984672.0, "step": 20859 }, { "epoch": 2.653606411398041, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.347006797790527, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8554826974868774, "num_tokens": 796022599.0, "step": 20860 }, { "epoch": 2.6537336216766314, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.362060546875, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8730131387710571, "num_tokens": 796057798.0, "step": 20861 }, { "epoch": 2.653860831955222, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.301671981811523, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8745380640029907, "num_tokens": 796094928.0, "step": 20862 }, { "epoch": 2.6539880422338125, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.313405990600586, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8761987090110779, "num_tokens": 796132614.0, "step": 20863 }, { "epoch": 2.654115252512403, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.310208320617676, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8736951351165771, "num_tokens": 796171653.0, "step": 20864 }, { "epoch": 2.6542424627909935, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.348179817199707, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8691217303276062, "num_tokens": 796215218.0, "step": 20865 }, { "epoch": 2.654369673069584, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.359427452087402, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8864484429359436, "num_tokens": 796249165.0, "step": 20866 }, { "epoch": 2.6544968833481746, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.371316909790039, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8724942207336426, "num_tokens": 796283878.0, "step": 20867 }, { "epoch": 2.654624093626765, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.280385971069336, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8622564673423767, "num_tokens": 796330760.0, "step": 20868 }, { "epoch": 2.6547513039053556, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.369606971740723, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.863570511341095, "num_tokens": 796371547.0, "step": 20869 }, { "epoch": 2.654878514183946, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.3143892288208, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8597699999809265, "num_tokens": 796409893.0, "step": 20870 }, { "epoch": 2.6550057244625367, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.383477210998535, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8595137000083923, "num_tokens": 796442493.0, "step": 20871 }, { "epoch": 2.655132934741127, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.365829467773438, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.877184271812439, "num_tokens": 796480853.0, "step": 20872 }, { "epoch": 2.6552601450197177, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.353792190551758, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8806904554367065, "num_tokens": 796520057.0, "step": 20873 }, { "epoch": 2.6553873552983083, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.336023330688477, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8788944482803345, "num_tokens": 796549779.0, "step": 20874 }, { "epoch": 2.6555145655768984, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.365476608276367, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8799921274185181, "num_tokens": 796590923.0, "step": 20875 }, { "epoch": 2.6556417758554893, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.356887817382812, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8703615665435791, "num_tokens": 796629455.0, "step": 20876 }, { "epoch": 2.6557689861340794, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.3690767288208, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8708704710006714, "num_tokens": 796669950.0, "step": 20877 }, { "epoch": 2.6558961964126704, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.353399276733398, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8631356358528137, "num_tokens": 796712263.0, "step": 20878 }, { "epoch": 2.6560234066912605, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.344770431518555, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.874039888381958, "num_tokens": 796753025.0, "step": 20879 }, { "epoch": 2.6561506169698514, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.274823188781738, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.886483371257782, "num_tokens": 796788810.0, "step": 20880 }, { "epoch": 2.6562778272484415, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.359865188598633, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8769731521606445, "num_tokens": 796826275.0, "step": 20881 }, { "epoch": 2.656405037527032, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.298334121704102, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8692190647125244, "num_tokens": 796859587.0, "step": 20882 }, { "epoch": 2.6565322478056226, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.373811721801758, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8804795145988464, "num_tokens": 796899440.0, "step": 20883 }, { "epoch": 2.656659458084213, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.305827140808105, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8648939728736877, "num_tokens": 796938703.0, "step": 20884 }, { "epoch": 2.6567866683628036, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.426389694213867, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8611066341400146, "num_tokens": 796982287.0, "step": 20885 }, { "epoch": 2.656913878641394, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.429671287536621, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8769305348396301, "num_tokens": 797017640.0, "step": 20886 }, { "epoch": 2.6570410889199847, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.335018157958984, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8597579002380371, "num_tokens": 797054050.0, "step": 20887 }, { "epoch": 2.6571682991985752, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.415529251098633, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8788634538650513, "num_tokens": 797092429.0, "step": 20888 }, { "epoch": 2.6572955094771658, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.273022651672363, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8710939884185791, "num_tokens": 797125646.0, "step": 20889 }, { "epoch": 2.6574227197557563, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.408586502075195, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8766950368881226, "num_tokens": 797164560.0, "step": 20890 }, { "epoch": 2.657549930034347, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.328124046325684, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8730908632278442, "num_tokens": 797194168.0, "step": 20891 }, { "epoch": 2.6576771403129373, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.342748641967773, "learning_rate": 1e-06, "loss": 0.343, "mean_token_accuracy": 0.893236517906189, "num_tokens": 797227045.0, "step": 20892 }, { "epoch": 2.657804350591528, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.327005386352539, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.889075517654419, "num_tokens": 797267607.0, "step": 20893 }, { "epoch": 2.6579315608701184, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.38818359375, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8636481165885925, "num_tokens": 797309320.0, "step": 20894 }, { "epoch": 2.658058771148709, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.294554710388184, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8714833855628967, "num_tokens": 797348258.0, "step": 20895 }, { "epoch": 2.6581859814272994, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.314334869384766, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.873874306678772, "num_tokens": 797391563.0, "step": 20896 }, { "epoch": 2.65831319170589, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.317084312438965, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8831301331520081, "num_tokens": 797428314.0, "step": 20897 }, { "epoch": 2.6584404019844805, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.391220092773438, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8534694910049438, "num_tokens": 797464480.0, "step": 20898 }, { "epoch": 2.658567612263071, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.365130424499512, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8756019473075867, "num_tokens": 797504135.0, "step": 20899 }, { "epoch": 2.658694822541661, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.338306427001953, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.877192497253418, "num_tokens": 797544222.0, "step": 20900 }, { "epoch": 2.658822032820252, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.27716064453125, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.879845917224884, "num_tokens": 797583695.0, "step": 20901 }, { "epoch": 2.658949243098842, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.312572479248047, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8605462312698364, "num_tokens": 797625660.0, "step": 20902 }, { "epoch": 2.659076453377433, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.34511947631836, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8740915656089783, "num_tokens": 797665807.0, "step": 20903 }, { "epoch": 2.6592036636560232, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.259596824645996, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8578318357467651, "num_tokens": 797707935.0, "step": 20904 }, { "epoch": 2.6593308739346138, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.339932441711426, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8678919076919556, "num_tokens": 797745502.0, "step": 20905 }, { "epoch": 2.6594580842132043, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.323497772216797, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8716228604316711, "num_tokens": 797779412.0, "step": 20906 }, { "epoch": 2.659585294491795, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.3433256149292, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8772329092025757, "num_tokens": 797816677.0, "step": 20907 }, { "epoch": 2.6597125047703853, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.258856773376465, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8570832014083862, "num_tokens": 797854260.0, "step": 20908 }, { "epoch": 2.659839715048976, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.367513656616211, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8614226579666138, "num_tokens": 797892349.0, "step": 20909 }, { "epoch": 2.6599669253275664, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.297262191772461, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.878116250038147, "num_tokens": 797928028.0, "step": 20910 }, { "epoch": 2.660094135606157, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.393414497375488, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8591421842575073, "num_tokens": 797963885.0, "step": 20911 }, { "epoch": 2.6602213458847475, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.309981346130371, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8608594536781311, "num_tokens": 798009842.0, "step": 20912 }, { "epoch": 2.660348556163338, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.344268798828125, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8860583901405334, "num_tokens": 798043207.0, "step": 20913 }, { "epoch": 2.6604757664419285, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.336942672729492, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8897360563278198, "num_tokens": 798080887.0, "step": 20914 }, { "epoch": 2.660602976720519, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.365221977233887, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8835867643356323, "num_tokens": 798126393.0, "step": 20915 }, { "epoch": 2.6607301869991096, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.336901664733887, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8896481990814209, "num_tokens": 798166099.0, "step": 20916 }, { "epoch": 2.6608573972777, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.396384239196777, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8681062459945679, "num_tokens": 798202285.0, "step": 20917 }, { "epoch": 2.6609846075562906, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.34249496459961, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8811503052711487, "num_tokens": 798243463.0, "step": 20918 }, { "epoch": 2.661111817834881, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.379382133483887, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.876495361328125, "num_tokens": 798275507.0, "step": 20919 }, { "epoch": 2.6612390281134717, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.370022773742676, "learning_rate": 1e-06, "loss": 0.3407, "mean_token_accuracy": 0.8922395706176758, "num_tokens": 798310536.0, "step": 20920 }, { "epoch": 2.661366238392062, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.413651466369629, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8690969944000244, "num_tokens": 798347874.0, "step": 20921 }, { "epoch": 2.6614934486706527, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.378098487854004, "learning_rate": 1e-06, "loss": 0.3252, "mean_token_accuracy": 0.8975427150726318, "num_tokens": 798381865.0, "step": 20922 }, { "epoch": 2.661620658949243, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.307025909423828, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8735477924346924, "num_tokens": 798423378.0, "step": 20923 }, { "epoch": 2.661747869227834, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.368350982666016, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8809279203414917, "num_tokens": 798460278.0, "step": 20924 }, { "epoch": 2.661875079506424, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.402360916137695, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8750635385513306, "num_tokens": 798496622.0, "step": 20925 }, { "epoch": 2.662002289785015, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.366759300231934, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8697682023048401, "num_tokens": 798533061.0, "step": 20926 }, { "epoch": 2.662129500063605, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.391035079956055, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8669660091400146, "num_tokens": 798573689.0, "step": 20927 }, { "epoch": 2.662256710342196, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.327606201171875, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8830905556678772, "num_tokens": 798614058.0, "step": 20928 }, { "epoch": 2.662383920620786, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.328349113464355, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8640332221984863, "num_tokens": 798657769.0, "step": 20929 }, { "epoch": 2.6625111308993765, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.324285507202148, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8775798082351685, "num_tokens": 798701084.0, "step": 20930 }, { "epoch": 2.662638341177967, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.373339653015137, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8801514506340027, "num_tokens": 798738029.0, "step": 20931 }, { "epoch": 2.6627655514565576, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.35236644744873, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8697309494018555, "num_tokens": 798773537.0, "step": 20932 }, { "epoch": 2.662892761735148, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.324606895446777, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8712972402572632, "num_tokens": 798810397.0, "step": 20933 }, { "epoch": 2.6630199720137386, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.339266777038574, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8711477518081665, "num_tokens": 798850796.0, "step": 20934 }, { "epoch": 2.663147182292329, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.333333015441895, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8737387657165527, "num_tokens": 798892379.0, "step": 20935 }, { "epoch": 2.6632743925709197, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.353349685668945, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8623101115226746, "num_tokens": 798929444.0, "step": 20936 }, { "epoch": 2.66340160284951, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.39847183227539, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8805183172225952, "num_tokens": 798964086.0, "step": 20937 }, { "epoch": 2.6635288131281007, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.332679748535156, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8818209171295166, "num_tokens": 799007408.0, "step": 20938 }, { "epoch": 2.6636560234066913, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.409854888916016, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8788735866546631, "num_tokens": 799045930.0, "step": 20939 }, { "epoch": 2.663783233685282, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.364895820617676, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8840034604072571, "num_tokens": 799083568.0, "step": 20940 }, { "epoch": 2.6639104439638723, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.405994415283203, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8708560466766357, "num_tokens": 799123681.0, "step": 20941 }, { "epoch": 2.664037654242463, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.420958518981934, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.870611310005188, "num_tokens": 799164623.0, "step": 20942 }, { "epoch": 2.6641648645210534, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.381295204162598, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8737908005714417, "num_tokens": 799200408.0, "step": 20943 }, { "epoch": 2.664292074799644, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.303524017333984, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8631158471107483, "num_tokens": 799245106.0, "step": 20944 }, { "epoch": 2.6644192850782344, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.345786094665527, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8817861080169678, "num_tokens": 799290429.0, "step": 20945 }, { "epoch": 2.664546495356825, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.428595542907715, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8727737665176392, "num_tokens": 799322951.0, "step": 20946 }, { "epoch": 2.6646737056354155, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.3829927444458, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8708568811416626, "num_tokens": 799365498.0, "step": 20947 }, { "epoch": 2.6648009159140056, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.286463737487793, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8679283857345581, "num_tokens": 799405894.0, "step": 20948 }, { "epoch": 2.6649281261925966, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.464147567749023, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.873233437538147, "num_tokens": 799439750.0, "step": 20949 }, { "epoch": 2.6650553364711866, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.308313369750977, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8651612997055054, "num_tokens": 799475170.0, "step": 20950 }, { "epoch": 2.6651825467497776, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.391266822814941, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8618900775909424, "num_tokens": 799514231.0, "step": 20951 }, { "epoch": 2.6653097570283677, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.301798820495605, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8751809597015381, "num_tokens": 799549512.0, "step": 20952 }, { "epoch": 2.6654369673069587, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.337345123291016, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8696256875991821, "num_tokens": 799592144.0, "step": 20953 }, { "epoch": 2.6655641775855488, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.325597763061523, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8595739603042603, "num_tokens": 799634779.0, "step": 20954 }, { "epoch": 2.6656913878641393, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.293399810791016, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8856470584869385, "num_tokens": 799677548.0, "step": 20955 }, { "epoch": 2.66581859814273, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.398398399353027, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8568687438964844, "num_tokens": 799715679.0, "step": 20956 }, { "epoch": 2.6659458084213203, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.314335823059082, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8672741651535034, "num_tokens": 799755701.0, "step": 20957 }, { "epoch": 2.666073018699911, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.394152641296387, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8687794804573059, "num_tokens": 799795151.0, "step": 20958 }, { "epoch": 2.6662002289785014, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.359158515930176, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8803989887237549, "num_tokens": 799832249.0, "step": 20959 }, { "epoch": 2.666327439257092, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.341254234313965, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8537977933883667, "num_tokens": 799867762.0, "step": 20960 }, { "epoch": 2.6664546495356825, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.360492706298828, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8706612586975098, "num_tokens": 799909012.0, "step": 20961 }, { "epoch": 2.666581859814273, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.333081245422363, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8708648681640625, "num_tokens": 799950316.0, "step": 20962 }, { "epoch": 2.6667090700928635, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.38404369354248, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8883733749389648, "num_tokens": 799990037.0, "step": 20963 }, { "epoch": 2.666836280371454, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.440860748291016, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8826199769973755, "num_tokens": 800023935.0, "step": 20964 }, { "epoch": 2.6669634906500446, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.350685119628906, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8793648481369019, "num_tokens": 800066357.0, "step": 20965 }, { "epoch": 2.667090700928635, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.493680000305176, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8602193593978882, "num_tokens": 800100667.0, "step": 20966 }, { "epoch": 2.6672179112072256, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.369108200073242, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8777283430099487, "num_tokens": 800135101.0, "step": 20967 }, { "epoch": 2.667345121485816, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.424246788024902, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8882782459259033, "num_tokens": 800176308.0, "step": 20968 }, { "epoch": 2.6674723317644067, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.389654159545898, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8849743604660034, "num_tokens": 800205585.0, "step": 20969 }, { "epoch": 2.667599542042997, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.358662605285645, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8732962608337402, "num_tokens": 800241400.0, "step": 20970 }, { "epoch": 2.6677267523215877, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.432518005371094, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8837440013885498, "num_tokens": 800279060.0, "step": 20971 }, { "epoch": 2.6678539626001783, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.370445251464844, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8829174041748047, "num_tokens": 800318606.0, "step": 20972 }, { "epoch": 2.6679811728787683, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.463935852050781, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8730000853538513, "num_tokens": 800356795.0, "step": 20973 }, { "epoch": 2.6681083831573593, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.326783180236816, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8818346858024597, "num_tokens": 800393786.0, "step": 20974 }, { "epoch": 2.6682355934359494, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.360452651977539, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8719953894615173, "num_tokens": 800440153.0, "step": 20975 }, { "epoch": 2.6683628037145404, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.417571067810059, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.856479287147522, "num_tokens": 800476343.0, "step": 20976 }, { "epoch": 2.6684900139931305, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.309244155883789, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8863649368286133, "num_tokens": 800514227.0, "step": 20977 }, { "epoch": 2.668617224271721, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.76672649383545, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8725224137306213, "num_tokens": 800552357.0, "step": 20978 }, { "epoch": 2.6687444345503115, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.283944129943848, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8811428546905518, "num_tokens": 800587624.0, "step": 20979 }, { "epoch": 2.668871644828902, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.347895622253418, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8678840398788452, "num_tokens": 800631183.0, "step": 20980 }, { "epoch": 2.6689988551074926, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.36991024017334, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8780585527420044, "num_tokens": 800670334.0, "step": 20981 }, { "epoch": 2.669126065386083, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.355598449707031, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8705457448959351, "num_tokens": 800703500.0, "step": 20982 }, { "epoch": 2.6692532756646736, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.363290786743164, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8670667409896851, "num_tokens": 800742904.0, "step": 20983 }, { "epoch": 2.669380485943264, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.386198043823242, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8718321323394775, "num_tokens": 800777556.0, "step": 20984 }, { "epoch": 2.6695076962218547, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.35558032989502, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8644939661026001, "num_tokens": 800817532.0, "step": 20985 }, { "epoch": 2.669634906500445, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.376052856445312, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8830649852752686, "num_tokens": 800856567.0, "step": 20986 }, { "epoch": 2.6697621167790357, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.376423835754395, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8521487712860107, "num_tokens": 800901234.0, "step": 20987 }, { "epoch": 2.6698893270576263, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.336722373962402, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8711413145065308, "num_tokens": 800942647.0, "step": 20988 }, { "epoch": 2.670016537336217, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.332931518554688, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8803625106811523, "num_tokens": 800980194.0, "step": 20989 }, { "epoch": 2.6701437476148073, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.443702697753906, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8779672384262085, "num_tokens": 801016978.0, "step": 20990 }, { "epoch": 2.670270957893398, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.318157196044922, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8600081205368042, "num_tokens": 801051921.0, "step": 20991 }, { "epoch": 2.6703981681719884, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.401104927062988, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.848829984664917, "num_tokens": 801087189.0, "step": 20992 }, { "epoch": 2.670525378450579, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.403511047363281, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8793375492095947, "num_tokens": 801121574.0, "step": 20993 }, { "epoch": 2.6706525887291694, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.401196479797363, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8786986470222473, "num_tokens": 801160497.0, "step": 20994 }, { "epoch": 2.67077979900776, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.396862030029297, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.87459796667099, "num_tokens": 801198556.0, "step": 20995 }, { "epoch": 2.6709070092863505, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.282943725585938, "learning_rate": 1e-06, "loss": 0.3194, "mean_token_accuracy": 0.8988028764724731, "num_tokens": 801230616.0, "step": 20996 }, { "epoch": 2.671034219564941, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.375101089477539, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8720240592956543, "num_tokens": 801266582.0, "step": 20997 }, { "epoch": 2.671161429843531, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.313185691833496, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8855471611022949, "num_tokens": 801305127.0, "step": 20998 }, { "epoch": 2.671288640122122, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2686185836792, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8786622881889343, "num_tokens": 801348284.0, "step": 20999 }, { "epoch": 2.671415850400712, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.394441604614258, "learning_rate": 1e-06, "loss": 0.3365, "mean_token_accuracy": 0.8929948806762695, "num_tokens": 801385298.0, "step": 21000 }, { "epoch": 2.671543060679303, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.388932228088379, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8658491373062134, "num_tokens": 801417020.0, "step": 21001 }, { "epoch": 2.6716702709578932, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.332524299621582, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8666085004806519, "num_tokens": 801455328.0, "step": 21002 }, { "epoch": 2.6717974812364838, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.401819229125977, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8842566013336182, "num_tokens": 801494342.0, "step": 21003 }, { "epoch": 2.6719246915150743, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.304535865783691, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8762096166610718, "num_tokens": 801528182.0, "step": 21004 }, { "epoch": 2.672051901793665, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.404894828796387, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8741277456283569, "num_tokens": 801569424.0, "step": 21005 }, { "epoch": 2.6721791120722553, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.366353034973145, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8714852929115295, "num_tokens": 801608809.0, "step": 21006 }, { "epoch": 2.672306322350846, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2948579788208, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8649345636367798, "num_tokens": 801651910.0, "step": 21007 }, { "epoch": 2.6724335326294364, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.36776351928711, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8730791211128235, "num_tokens": 801692961.0, "step": 21008 }, { "epoch": 2.672560742908027, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.327739715576172, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8620679378509521, "num_tokens": 801731278.0, "step": 21009 }, { "epoch": 2.6726879531866174, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.360875129699707, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.870305061340332, "num_tokens": 801769562.0, "step": 21010 }, { "epoch": 2.672815163465208, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.333147048950195, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8695858716964722, "num_tokens": 801811046.0, "step": 21011 }, { "epoch": 2.6729423737437985, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.37342643737793, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8505821228027344, "num_tokens": 801847833.0, "step": 21012 }, { "epoch": 2.673069584022389, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.294353485107422, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8684008121490479, "num_tokens": 801887819.0, "step": 21013 }, { "epoch": 2.6731967943009796, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.260224342346191, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8847771286964417, "num_tokens": 801922535.0, "step": 21014 }, { "epoch": 2.67332400457957, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.407208442687988, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8784431219100952, "num_tokens": 801958025.0, "step": 21015 }, { "epoch": 2.6734512148581606, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.297162055969238, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8778291940689087, "num_tokens": 801991291.0, "step": 21016 }, { "epoch": 2.673578425136751, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.389859199523926, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.875793993473053, "num_tokens": 802024952.0, "step": 21017 }, { "epoch": 2.6737056354153417, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.384737014770508, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8730268478393555, "num_tokens": 802065741.0, "step": 21018 }, { "epoch": 2.673832845693932, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.271072387695312, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8725063800811768, "num_tokens": 802107153.0, "step": 21019 }, { "epoch": 2.6739600559725227, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.332895278930664, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.879015326499939, "num_tokens": 802148177.0, "step": 21020 }, { "epoch": 2.674087266251113, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.2855806350708, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8753064870834351, "num_tokens": 802186144.0, "step": 21021 }, { "epoch": 2.674214476529704, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.3125581741333, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8874413967132568, "num_tokens": 802224054.0, "step": 21022 }, { "epoch": 2.674341686808294, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.353353500366211, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8577395677566528, "num_tokens": 802264483.0, "step": 21023 }, { "epoch": 2.674468897086885, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.307517051696777, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8816788792610168, "num_tokens": 802302933.0, "step": 21024 }, { "epoch": 2.674596107365475, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.325839042663574, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8671367764472961, "num_tokens": 802347354.0, "step": 21025 }, { "epoch": 2.674723317644066, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.412858009338379, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.882564127445221, "num_tokens": 802392159.0, "step": 21026 }, { "epoch": 2.674850527922656, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.357795715332031, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.877612292766571, "num_tokens": 802432629.0, "step": 21027 }, { "epoch": 2.6749777382012465, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.431535720825195, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8828190565109253, "num_tokens": 802471776.0, "step": 21028 }, { "epoch": 2.675104948479837, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.35957145690918, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.867563009262085, "num_tokens": 802507314.0, "step": 21029 }, { "epoch": 2.6752321587584276, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.29578971862793, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8818151950836182, "num_tokens": 802550587.0, "step": 21030 }, { "epoch": 2.675359369037018, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.389704704284668, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8750081062316895, "num_tokens": 802591544.0, "step": 21031 }, { "epoch": 2.6754865793156086, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.328995704650879, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8808698654174805, "num_tokens": 802627336.0, "step": 21032 }, { "epoch": 2.675613789594199, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.354979515075684, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8731375336647034, "num_tokens": 802668358.0, "step": 21033 }, { "epoch": 2.6757409998727897, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.319509506225586, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8808459043502808, "num_tokens": 802708023.0, "step": 21034 }, { "epoch": 2.67586821015138, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.416255950927734, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8941348791122437, "num_tokens": 802745856.0, "step": 21035 }, { "epoch": 2.6759954204299707, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.310030937194824, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.88630211353302, "num_tokens": 802779032.0, "step": 21036 }, { "epoch": 2.6761226307085613, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.351607322692871, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8743917942047119, "num_tokens": 802815736.0, "step": 21037 }, { "epoch": 2.676249840987152, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.299507141113281, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8704394698143005, "num_tokens": 802854133.0, "step": 21038 }, { "epoch": 2.6763770512657423, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.283852577209473, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8812787532806396, "num_tokens": 802892969.0, "step": 21039 }, { "epoch": 2.676504261544333, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.328371047973633, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.889293372631073, "num_tokens": 802936998.0, "step": 21040 }, { "epoch": 2.6766314718229234, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.26624870300293, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8821114301681519, "num_tokens": 802979040.0, "step": 21041 }, { "epoch": 2.676758682101514, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.339634895324707, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8855463266372681, "num_tokens": 803016516.0, "step": 21042 }, { "epoch": 2.6768858923801044, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.3212251663208, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8728289604187012, "num_tokens": 803055566.0, "step": 21043 }, { "epoch": 2.677013102658695, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.340093612670898, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8761545419692993, "num_tokens": 803093683.0, "step": 21044 }, { "epoch": 2.6771403129372855, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.293170928955078, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.883515477180481, "num_tokens": 803133691.0, "step": 21045 }, { "epoch": 2.6772675232158756, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.349472999572754, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8830709457397461, "num_tokens": 803169870.0, "step": 21046 }, { "epoch": 2.6773947334944665, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.416983604431152, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8665310144424438, "num_tokens": 803209038.0, "step": 21047 }, { "epoch": 2.6775219437730566, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.305886268615723, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8609265089035034, "num_tokens": 803239725.0, "step": 21048 }, { "epoch": 2.6776491540516476, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.38998794555664, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8730604648590088, "num_tokens": 803273580.0, "step": 21049 }, { "epoch": 2.6777763643302377, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.327174186706543, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8802666664123535, "num_tokens": 803313583.0, "step": 21050 }, { "epoch": 2.6779035746088287, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.347705841064453, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8826002478599548, "num_tokens": 803347517.0, "step": 21051 }, { "epoch": 2.6780307848874187, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.444539070129395, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8617303371429443, "num_tokens": 803388183.0, "step": 21052 }, { "epoch": 2.6781579951660093, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.46453857421875, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8738662004470825, "num_tokens": 803428377.0, "step": 21053 }, { "epoch": 2.6782852054446, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.358963966369629, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8826843500137329, "num_tokens": 803460930.0, "step": 21054 }, { "epoch": 2.6784124157231903, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.402298927307129, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8827518224716187, "num_tokens": 803497971.0, "step": 21055 }, { "epoch": 2.678539626001781, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.32457447052002, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8894712924957275, "num_tokens": 803535514.0, "step": 21056 }, { "epoch": 2.6786668362803714, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.402997016906738, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8705072402954102, "num_tokens": 803580767.0, "step": 21057 }, { "epoch": 2.678794046558962, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.316065788269043, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8775732517242432, "num_tokens": 803619168.0, "step": 21058 }, { "epoch": 2.6789212568375524, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.339048385620117, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8606668710708618, "num_tokens": 803654785.0, "step": 21059 }, { "epoch": 2.679048467116143, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.293149948120117, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8830355405807495, "num_tokens": 803692505.0, "step": 21060 }, { "epoch": 2.6791756773947335, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.885214805603027, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8821313381195068, "num_tokens": 803732592.0, "step": 21061 }, { "epoch": 2.679302887673324, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.3381986618042, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8630252480506897, "num_tokens": 803767601.0, "step": 21062 }, { "epoch": 2.6794300979519146, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.261123657226562, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8669809103012085, "num_tokens": 803804329.0, "step": 21063 }, { "epoch": 2.679557308230505, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.449111938476562, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8695578575134277, "num_tokens": 803846182.0, "step": 21064 }, { "epoch": 2.6796845185090956, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.324812889099121, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.870208203792572, "num_tokens": 803883561.0, "step": 21065 }, { "epoch": 2.679811728787686, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.344969749450684, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8934438228607178, "num_tokens": 803916860.0, "step": 21066 }, { "epoch": 2.6799389390662767, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.307265281677246, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8757172226905823, "num_tokens": 803951949.0, "step": 21067 }, { "epoch": 2.680066149344867, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.393564224243164, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8707088232040405, "num_tokens": 803992094.0, "step": 21068 }, { "epoch": 2.6801933596234577, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.419499397277832, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8752235770225525, "num_tokens": 804034058.0, "step": 21069 }, { "epoch": 2.6803205699020483, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.35435676574707, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8665121793746948, "num_tokens": 804072722.0, "step": 21070 }, { "epoch": 2.6804477801806383, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.369473457336426, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8803477883338928, "num_tokens": 804114253.0, "step": 21071 }, { "epoch": 2.6805749904592293, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.39665699005127, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8775599002838135, "num_tokens": 804154877.0, "step": 21072 }, { "epoch": 2.6807022007378194, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.41943645477295, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8698883056640625, "num_tokens": 804191060.0, "step": 21073 }, { "epoch": 2.6808294110164104, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.406030654907227, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8667643070220947, "num_tokens": 804231574.0, "step": 21074 }, { "epoch": 2.6809566212950005, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.432169914245605, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8532958030700684, "num_tokens": 804264027.0, "step": 21075 }, { "epoch": 2.681083831573591, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.398395538330078, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8761976361274719, "num_tokens": 804302906.0, "step": 21076 }, { "epoch": 2.6812110418521815, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.395434379577637, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8703314065933228, "num_tokens": 804339107.0, "step": 21077 }, { "epoch": 2.681338252130772, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.41491413116455, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8800517916679382, "num_tokens": 804374292.0, "step": 21078 }, { "epoch": 2.6814654624093626, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.387535095214844, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8764265775680542, "num_tokens": 804420287.0, "step": 21079 }, { "epoch": 2.681592672687953, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.433533668518066, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8795289397239685, "num_tokens": 804453977.0, "step": 21080 }, { "epoch": 2.6817198829665436, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.411998748779297, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8590938448905945, "num_tokens": 804494958.0, "step": 21081 }, { "epoch": 2.681847093245134, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.435494422912598, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8553655743598938, "num_tokens": 804532209.0, "step": 21082 }, { "epoch": 2.6819743035237247, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.419957160949707, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8771606683731079, "num_tokens": 804572937.0, "step": 21083 }, { "epoch": 2.682101513802315, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.335323333740234, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8747503757476807, "num_tokens": 804610263.0, "step": 21084 }, { "epoch": 2.6822287240809057, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.375508308410645, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8785617351531982, "num_tokens": 804645516.0, "step": 21085 }, { "epoch": 2.6823559343594963, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.304964065551758, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8854305744171143, "num_tokens": 804684540.0, "step": 21086 }, { "epoch": 2.682483144638087, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.356441497802734, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8764384388923645, "num_tokens": 804723030.0, "step": 21087 }, { "epoch": 2.6826103549166773, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.332751274108887, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8837341666221619, "num_tokens": 804759202.0, "step": 21088 }, { "epoch": 2.682737565195268, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.443239212036133, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8713264465332031, "num_tokens": 804803199.0, "step": 21089 }, { "epoch": 2.6828647754738584, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.337642669677734, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8722522258758545, "num_tokens": 804835675.0, "step": 21090 }, { "epoch": 2.682991985752449, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.4918794631958, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8618730306625366, "num_tokens": 804876583.0, "step": 21091 }, { "epoch": 2.6831191960310394, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.37240219116211, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.865349531173706, "num_tokens": 804906110.0, "step": 21092 }, { "epoch": 2.68324640630963, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.537299156188965, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8655478954315186, "num_tokens": 804942196.0, "step": 21093 }, { "epoch": 2.6833736165882205, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.382750511169434, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8743642568588257, "num_tokens": 804978760.0, "step": 21094 }, { "epoch": 2.683500826866811, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.470800399780273, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8671742677688599, "num_tokens": 805014212.0, "step": 21095 }, { "epoch": 2.683628037145401, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.38007926940918, "learning_rate": 1e-06, "loss": 0.3277, "mean_token_accuracy": 0.8965734243392944, "num_tokens": 805050528.0, "step": 21096 }, { "epoch": 2.683755247423992, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.356021881103516, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8746896982192993, "num_tokens": 805087355.0, "step": 21097 }, { "epoch": 2.683882457702582, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.436868667602539, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8603833317756653, "num_tokens": 805117837.0, "step": 21098 }, { "epoch": 2.684009667981173, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.376301765441895, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8735880851745605, "num_tokens": 805157734.0, "step": 21099 }, { "epoch": 2.684136878259763, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.292308807373047, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8812801241874695, "num_tokens": 805190657.0, "step": 21100 }, { "epoch": 2.6842640885383537, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.32642650604248, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8679906129837036, "num_tokens": 805229654.0, "step": 21101 }, { "epoch": 2.6843912988169443, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.358749389648438, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8744056224822998, "num_tokens": 805269199.0, "step": 21102 }, { "epoch": 2.684518509095535, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.433611869812012, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8690727949142456, "num_tokens": 805308918.0, "step": 21103 }, { "epoch": 2.6846457193741253, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.371052742004395, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8709677457809448, "num_tokens": 805345014.0, "step": 21104 }, { "epoch": 2.684772929652716, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.324295043945312, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.887642502784729, "num_tokens": 805383674.0, "step": 21105 }, { "epoch": 2.6849001399313064, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.451308250427246, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8787702322006226, "num_tokens": 805419765.0, "step": 21106 }, { "epoch": 2.685027350209897, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.312896728515625, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8750605583190918, "num_tokens": 805461346.0, "step": 21107 }, { "epoch": 2.6851545604884874, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.37592601776123, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8823071718215942, "num_tokens": 805494519.0, "step": 21108 }, { "epoch": 2.685281770767078, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.362485885620117, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8632515668869019, "num_tokens": 805530542.0, "step": 21109 }, { "epoch": 2.6854089810456685, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.358536720275879, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8732477426528931, "num_tokens": 805564531.0, "step": 21110 }, { "epoch": 2.685536191324259, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.429183006286621, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.871098518371582, "num_tokens": 805606011.0, "step": 21111 }, { "epoch": 2.6856634016028496, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.274616241455078, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8684419393539429, "num_tokens": 805643078.0, "step": 21112 }, { "epoch": 2.68579061188144, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.33459186553955, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8773953914642334, "num_tokens": 805678554.0, "step": 21113 }, { "epoch": 2.6859178221600306, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.351790428161621, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8839354515075684, "num_tokens": 805713736.0, "step": 21114 }, { "epoch": 2.686045032438621, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.322574615478516, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8683570623397827, "num_tokens": 805751798.0, "step": 21115 }, { "epoch": 2.6861722427172117, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.277848243713379, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8812150359153748, "num_tokens": 805789205.0, "step": 21116 }, { "epoch": 2.686299452995802, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.33930492401123, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8773164749145508, "num_tokens": 805824910.0, "step": 21117 }, { "epoch": 2.6864266632743927, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.375411987304688, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8717837929725647, "num_tokens": 805861335.0, "step": 21118 }, { "epoch": 2.686553873552983, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.247088432312012, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8709293603897095, "num_tokens": 805901488.0, "step": 21119 }, { "epoch": 2.686681083831574, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.389021873474121, "learning_rate": 1e-06, "loss": 0.5054, "mean_token_accuracy": 0.8380835056304932, "num_tokens": 805941336.0, "step": 21120 }, { "epoch": 2.686808294110164, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.325950622558594, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8827250003814697, "num_tokens": 805974222.0, "step": 21121 }, { "epoch": 2.686935504388755, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.363941192626953, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8709704875946045, "num_tokens": 806014942.0, "step": 21122 }, { "epoch": 2.687062714667345, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.3016357421875, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8838609457015991, "num_tokens": 806048894.0, "step": 21123 }, { "epoch": 2.687189924945936, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.358216285705566, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8721506595611572, "num_tokens": 806089714.0, "step": 21124 }, { "epoch": 2.687317135224526, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.382864952087402, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8626891374588013, "num_tokens": 806128974.0, "step": 21125 }, { "epoch": 2.6874443455031165, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.32673454284668, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.874068021774292, "num_tokens": 806169857.0, "step": 21126 }, { "epoch": 2.687571555781707, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.33634090423584, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8752527832984924, "num_tokens": 806202826.0, "step": 21127 }, { "epoch": 2.6876987660602976, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.303179740905762, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.872291088104248, "num_tokens": 806242850.0, "step": 21128 }, { "epoch": 2.687825976338888, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.338214874267578, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8759466409683228, "num_tokens": 806279024.0, "step": 21129 }, { "epoch": 2.6879531866174786, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.378787994384766, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8597107529640198, "num_tokens": 806313669.0, "step": 21130 }, { "epoch": 2.688080396896069, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.28158187866211, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8768789768218994, "num_tokens": 806353128.0, "step": 21131 }, { "epoch": 2.6882076071746597, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.410112380981445, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8760783672332764, "num_tokens": 806385044.0, "step": 21132 }, { "epoch": 2.68833481745325, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.307117462158203, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8667985200881958, "num_tokens": 806422157.0, "step": 21133 }, { "epoch": 2.6884620277318407, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.335681915283203, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8739156723022461, "num_tokens": 806463180.0, "step": 21134 }, { "epoch": 2.6885892380104313, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.352621078491211, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8700549602508545, "num_tokens": 806502031.0, "step": 21135 }, { "epoch": 2.688716448289022, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.394960403442383, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8759428858757019, "num_tokens": 806539585.0, "step": 21136 }, { "epoch": 2.6888436585676123, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.37730884552002, "learning_rate": 1e-06, "loss": 0.3394, "mean_token_accuracy": 0.8930038809776306, "num_tokens": 806573147.0, "step": 21137 }, { "epoch": 2.688970868846203, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.345985412597656, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8860650062561035, "num_tokens": 806609785.0, "step": 21138 }, { "epoch": 2.6890980791247934, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.358205795288086, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8814316987991333, "num_tokens": 806647709.0, "step": 21139 }, { "epoch": 2.689225289403384, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.2861328125, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8778703212738037, "num_tokens": 806684448.0, "step": 21140 }, { "epoch": 2.6893524996819744, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.359834671020508, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8779630064964294, "num_tokens": 806725326.0, "step": 21141 }, { "epoch": 2.689479709960565, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.421432495117188, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8744582533836365, "num_tokens": 806759797.0, "step": 21142 }, { "epoch": 2.6896069202391555, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.3602294921875, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8807916641235352, "num_tokens": 806799986.0, "step": 21143 }, { "epoch": 2.6897341305177456, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.326682090759277, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8692129850387573, "num_tokens": 806843055.0, "step": 21144 }, { "epoch": 2.6898613407963365, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.231268882751465, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8885790109634399, "num_tokens": 806875897.0, "step": 21145 }, { "epoch": 2.6899885510749266, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.398749351501465, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8824455738067627, "num_tokens": 806915583.0, "step": 21146 }, { "epoch": 2.6901157613535176, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.373931884765625, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8769688606262207, "num_tokens": 806953851.0, "step": 21147 }, { "epoch": 2.6902429716321077, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.331634521484375, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8710319995880127, "num_tokens": 806988519.0, "step": 21148 }, { "epoch": 2.6903701819106987, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.352373123168945, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.877753734588623, "num_tokens": 807029612.0, "step": 21149 }, { "epoch": 2.6904973921892887, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.394357681274414, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.858827531337738, "num_tokens": 807064009.0, "step": 21150 }, { "epoch": 2.6906246024678793, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.351203918457031, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8721129894256592, "num_tokens": 807104694.0, "step": 21151 }, { "epoch": 2.69075181274647, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.395689010620117, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8537214994430542, "num_tokens": 807142500.0, "step": 21152 }, { "epoch": 2.6908790230250603, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.40257740020752, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8705432415008545, "num_tokens": 807172123.0, "step": 21153 }, { "epoch": 2.691006233303651, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.404531478881836, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8758659958839417, "num_tokens": 807206422.0, "step": 21154 }, { "epoch": 2.6911334435822414, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.323932647705078, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8602187037467957, "num_tokens": 807240867.0, "step": 21155 }, { "epoch": 2.691260653860832, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.433666229248047, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8831591606140137, "num_tokens": 807275838.0, "step": 21156 }, { "epoch": 2.6913878641394224, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.345726013183594, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.882940948009491, "num_tokens": 807315164.0, "step": 21157 }, { "epoch": 2.691515074418013, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.342220306396484, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8651232719421387, "num_tokens": 807358016.0, "step": 21158 }, { "epoch": 2.6916422846966035, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.34228515625, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.873035192489624, "num_tokens": 807402751.0, "step": 21159 }, { "epoch": 2.691769494975194, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.344093322753906, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8781179189682007, "num_tokens": 807442761.0, "step": 21160 }, { "epoch": 2.6918967052537845, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.33767318725586, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8818162679672241, "num_tokens": 807478550.0, "step": 21161 }, { "epoch": 2.692023915532375, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.300501823425293, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.882846474647522, "num_tokens": 807522756.0, "step": 21162 }, { "epoch": 2.6921511258109656, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.376496315002441, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8791719675064087, "num_tokens": 807561701.0, "step": 21163 }, { "epoch": 2.692278336089556, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.342044830322266, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8733294010162354, "num_tokens": 807604058.0, "step": 21164 }, { "epoch": 2.6924055463681467, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.387928009033203, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8798755407333374, "num_tokens": 807637344.0, "step": 21165 }, { "epoch": 2.692532756646737, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.40649700164795, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8684936761856079, "num_tokens": 807676867.0, "step": 21166 }, { "epoch": 2.6926599669253277, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.388550758361816, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8773218989372253, "num_tokens": 807707183.0, "step": 21167 }, { "epoch": 2.6927871772039182, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.397077560424805, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8664972186088562, "num_tokens": 807737617.0, "step": 21168 }, { "epoch": 2.6929143874825083, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.386073112487793, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8673032522201538, "num_tokens": 807774141.0, "step": 21169 }, { "epoch": 2.6930415977610993, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.413386344909668, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8803232908248901, "num_tokens": 807813375.0, "step": 21170 }, { "epoch": 2.6931688080396894, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.322819709777832, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8620790839195251, "num_tokens": 807852317.0, "step": 21171 }, { "epoch": 2.6932960183182804, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.388647079467773, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8758689165115356, "num_tokens": 807895266.0, "step": 21172 }, { "epoch": 2.6934232285968704, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.261544227600098, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.876473069190979, "num_tokens": 807938600.0, "step": 21173 }, { "epoch": 2.693550438875461, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.384489059448242, "learning_rate": 1e-06, "loss": 0.3359, "mean_token_accuracy": 0.8940763473510742, "num_tokens": 807974329.0, "step": 21174 }, { "epoch": 2.6936776491540515, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.382575988769531, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8856072425842285, "num_tokens": 808015666.0, "step": 21175 }, { "epoch": 2.693804859432642, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.388921737670898, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8848427534103394, "num_tokens": 808050176.0, "step": 21176 }, { "epoch": 2.6939320697112326, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.456486701965332, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.867661714553833, "num_tokens": 808091910.0, "step": 21177 }, { "epoch": 2.694059279989823, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.329437255859375, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8697589635848999, "num_tokens": 808128646.0, "step": 21178 }, { "epoch": 2.6941864902684136, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.418814659118652, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8786278367042542, "num_tokens": 808166544.0, "step": 21179 }, { "epoch": 2.694313700547004, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.386062622070312, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8643302917480469, "num_tokens": 808200051.0, "step": 21180 }, { "epoch": 2.6944409108255947, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.417535781860352, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8760006427764893, "num_tokens": 808239251.0, "step": 21181 }, { "epoch": 2.694568121104185, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.30380630493164, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8688845634460449, "num_tokens": 808280639.0, "step": 21182 }, { "epoch": 2.6946953313827757, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.388021469116211, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8805860877037048, "num_tokens": 808316141.0, "step": 21183 }, { "epoch": 2.6948225416613663, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.369049072265625, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8731356859207153, "num_tokens": 808359172.0, "step": 21184 }, { "epoch": 2.694949751939957, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.331910133361816, "learning_rate": 1e-06, "loss": 0.491, "mean_token_accuracy": 0.8469693064689636, "num_tokens": 808397823.0, "step": 21185 }, { "epoch": 2.6950769622185473, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.415431022644043, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8839874267578125, "num_tokens": 808435355.0, "step": 21186 }, { "epoch": 2.695204172497138, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.361652374267578, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8805490732192993, "num_tokens": 808470792.0, "step": 21187 }, { "epoch": 2.6953313827757284, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.375057220458984, "learning_rate": 1e-06, "loss": 0.3451, "mean_token_accuracy": 0.8913005590438843, "num_tokens": 808504591.0, "step": 21188 }, { "epoch": 2.695458593054319, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.39636516571045, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8852714896202087, "num_tokens": 808537167.0, "step": 21189 }, { "epoch": 2.6955858033329094, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.365437507629395, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8889527320861816, "num_tokens": 808572055.0, "step": 21190 }, { "epoch": 2.6957130136115, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.435357093811035, "learning_rate": 1e-06, "loss": 0.3503, "mean_token_accuracy": 0.8861615657806396, "num_tokens": 808605239.0, "step": 21191 }, { "epoch": 2.6958402238900905, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.380681037902832, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8526158332824707, "num_tokens": 808643227.0, "step": 21192 }, { "epoch": 2.695967434168681, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.400254249572754, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8690399527549744, "num_tokens": 808683855.0, "step": 21193 }, { "epoch": 2.696094644447271, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.350310325622559, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8801842927932739, "num_tokens": 808723467.0, "step": 21194 }, { "epoch": 2.696221854725862, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.386301040649414, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8638268113136292, "num_tokens": 808759805.0, "step": 21195 }, { "epoch": 2.696349065004452, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.382856369018555, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.858067512512207, "num_tokens": 808802024.0, "step": 21196 }, { "epoch": 2.696476275283043, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.339040756225586, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8665124177932739, "num_tokens": 808837086.0, "step": 21197 }, { "epoch": 2.696603485561633, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.41456127166748, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8734704256057739, "num_tokens": 808875261.0, "step": 21198 }, { "epoch": 2.6967306958402237, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.327275276184082, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8811414241790771, "num_tokens": 808913044.0, "step": 21199 }, { "epoch": 2.6968579061188143, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.317488670349121, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8819072246551514, "num_tokens": 808952374.0, "step": 21200 }, { "epoch": 2.696985116397405, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.353464126586914, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8748594522476196, "num_tokens": 808994397.0, "step": 21201 }, { "epoch": 2.6971123266759953, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.388057708740234, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8775043487548828, "num_tokens": 809036401.0, "step": 21202 }, { "epoch": 2.697239536954586, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.406641006469727, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8799731731414795, "num_tokens": 809073073.0, "step": 21203 }, { "epoch": 2.6973667472331764, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.47201156616211, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8744689226150513, "num_tokens": 809117348.0, "step": 21204 }, { "epoch": 2.697493957511767, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.342073440551758, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8872880935668945, "num_tokens": 809151357.0, "step": 21205 }, { "epoch": 2.6976211677903574, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.441067695617676, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8690418004989624, "num_tokens": 809189000.0, "step": 21206 }, { "epoch": 2.697748378068948, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.299242973327637, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8780355453491211, "num_tokens": 809227380.0, "step": 21207 }, { "epoch": 2.6978755883475385, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.438162803649902, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8743646144866943, "num_tokens": 809265131.0, "step": 21208 }, { "epoch": 2.698002798626129, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.424392700195312, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8738766312599182, "num_tokens": 809297190.0, "step": 21209 }, { "epoch": 2.6981300089047195, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.333328247070312, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8834102153778076, "num_tokens": 809336774.0, "step": 21210 }, { "epoch": 2.69825721918331, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.45107650756836, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8588781356811523, "num_tokens": 809376612.0, "step": 21211 }, { "epoch": 2.6983844294619006, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.340067863464355, "learning_rate": 1e-06, "loss": 0.3221, "mean_token_accuracy": 0.8971595168113708, "num_tokens": 809416058.0, "step": 21212 }, { "epoch": 2.698511639740491, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.499311447143555, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8849339485168457, "num_tokens": 809449580.0, "step": 21213 }, { "epoch": 2.6986388500190817, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.425564765930176, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8780626058578491, "num_tokens": 809484983.0, "step": 21214 }, { "epoch": 2.698766060297672, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.411725997924805, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8774809241294861, "num_tokens": 809521086.0, "step": 21215 }, { "epoch": 2.6988932705762627, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.427486419677734, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8724493384361267, "num_tokens": 809562475.0, "step": 21216 }, { "epoch": 2.699020480854853, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.340940475463867, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8612000942230225, "num_tokens": 809596686.0, "step": 21217 }, { "epoch": 2.6991476911334438, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.518293380737305, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.876046895980835, "num_tokens": 809636614.0, "step": 21218 }, { "epoch": 2.699274901412034, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.419520378112793, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8632314205169678, "num_tokens": 809674542.0, "step": 21219 }, { "epoch": 2.699402111690625, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.433247566223145, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8905736207962036, "num_tokens": 809716726.0, "step": 21220 }, { "epoch": 2.699529321969215, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.411065101623535, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8829452991485596, "num_tokens": 809756790.0, "step": 21221 }, { "epoch": 2.699656532247806, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.41061782836914, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8717581033706665, "num_tokens": 809794616.0, "step": 21222 }, { "epoch": 2.699783742526396, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.444599151611328, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8812405467033386, "num_tokens": 809836662.0, "step": 21223 }, { "epoch": 2.6999109528049865, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.457927703857422, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8722584843635559, "num_tokens": 809873812.0, "step": 21224 }, { "epoch": 2.700038163083577, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.41926097869873, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.87479567527771, "num_tokens": 809913139.0, "step": 21225 }, { "epoch": 2.7001653733621676, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.440492630004883, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8757927417755127, "num_tokens": 809947741.0, "step": 21226 }, { "epoch": 2.700292583640758, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.4330472946167, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8674759268760681, "num_tokens": 809987120.0, "step": 21227 }, { "epoch": 2.7004197939193486, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.396350860595703, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.869422435760498, "num_tokens": 810026972.0, "step": 21228 }, { "epoch": 2.700547004197939, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.441079139709473, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8741399049758911, "num_tokens": 810062767.0, "step": 21229 }, { "epoch": 2.7006742144765297, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.385507583618164, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8828210830688477, "num_tokens": 810096930.0, "step": 21230 }, { "epoch": 2.70080142475512, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.391745567321777, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8622848987579346, "num_tokens": 810134600.0, "step": 21231 }, { "epoch": 2.7009286350337107, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.455913543701172, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8768612146377563, "num_tokens": 810169535.0, "step": 21232 }, { "epoch": 2.7010558453123013, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.445243835449219, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.864452600479126, "num_tokens": 810205288.0, "step": 21233 }, { "epoch": 2.701183055590892, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.503328323364258, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8720400929450989, "num_tokens": 810242868.0, "step": 21234 }, { "epoch": 2.7013102658694823, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.359411239624023, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8628556728363037, "num_tokens": 810286539.0, "step": 21235 }, { "epoch": 2.701437476148073, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.428282737731934, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.878014326095581, "num_tokens": 810331598.0, "step": 21236 }, { "epoch": 2.7015646864266634, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.38988208770752, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8824225068092346, "num_tokens": 810369501.0, "step": 21237 }, { "epoch": 2.701691896705254, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.43111515045166, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8712519407272339, "num_tokens": 810410727.0, "step": 21238 }, { "epoch": 2.7018191069838444, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.422805786132812, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8670803904533386, "num_tokens": 810450315.0, "step": 21239 }, { "epoch": 2.701946317262435, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.433490753173828, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.867088258266449, "num_tokens": 810493972.0, "step": 21240 }, { "epoch": 2.7020735275410255, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.378233909606934, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8859543800354004, "num_tokens": 810526038.0, "step": 21241 }, { "epoch": 2.7022007378196156, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.394689559936523, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8787026405334473, "num_tokens": 810560772.0, "step": 21242 }, { "epoch": 2.7023279480982065, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.325319290161133, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.882485568523407, "num_tokens": 810598322.0, "step": 21243 }, { "epoch": 2.7024551583767966, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.371904373168945, "learning_rate": 1e-06, "loss": 0.3428, "mean_token_accuracy": 0.8940292596817017, "num_tokens": 810638652.0, "step": 21244 }, { "epoch": 2.7025823686553876, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.436227798461914, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8727908730506897, "num_tokens": 810680318.0, "step": 21245 }, { "epoch": 2.7027095789339777, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.436006546020508, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8789690732955933, "num_tokens": 810715001.0, "step": 21246 }, { "epoch": 2.7028367892125686, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.462379455566406, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8666613101959229, "num_tokens": 810748897.0, "step": 21247 }, { "epoch": 2.7029639994911587, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.389803886413574, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8688549995422363, "num_tokens": 810788954.0, "step": 21248 }, { "epoch": 2.7030912097697493, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.464556694030762, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8768492937088013, "num_tokens": 810828897.0, "step": 21249 }, { "epoch": 2.70321842004834, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.398063659667969, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.891116201877594, "num_tokens": 810866698.0, "step": 21250 }, { "epoch": 2.7033456303269303, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.536648750305176, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8676480650901794, "num_tokens": 810903140.0, "step": 21251 }, { "epoch": 2.703472840605521, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.414277076721191, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8789910078048706, "num_tokens": 810940664.0, "step": 21252 }, { "epoch": 2.7036000508841114, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.397216796875, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8727377653121948, "num_tokens": 810981818.0, "step": 21253 }, { "epoch": 2.703727261162702, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.420881271362305, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8738450407981873, "num_tokens": 811019897.0, "step": 21254 }, { "epoch": 2.7038544714412924, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.440025329589844, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.878040075302124, "num_tokens": 811056310.0, "step": 21255 }, { "epoch": 2.703981681719883, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.46207332611084, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8643800020217896, "num_tokens": 811091505.0, "step": 21256 }, { "epoch": 2.7041088919984735, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.443384170532227, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.884872317314148, "num_tokens": 811134861.0, "step": 21257 }, { "epoch": 2.704236102277064, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.410136222839355, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8666032552719116, "num_tokens": 811170001.0, "step": 21258 }, { "epoch": 2.7043633125556545, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.471022605895996, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8794178366661072, "num_tokens": 811204203.0, "step": 21259 }, { "epoch": 2.704490522834245, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.437888145446777, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8788172006607056, "num_tokens": 811244598.0, "step": 21260 }, { "epoch": 2.7046177331128356, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.427606582641602, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8845090866088867, "num_tokens": 811283352.0, "step": 21261 }, { "epoch": 2.704744943391426, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.353419303894043, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8634781837463379, "num_tokens": 811317605.0, "step": 21262 }, { "epoch": 2.7048721536700167, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.433721542358398, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8910865187644958, "num_tokens": 811352117.0, "step": 21263 }, { "epoch": 2.704999363948607, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.459368705749512, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8674684762954712, "num_tokens": 811386118.0, "step": 21264 }, { "epoch": 2.7051265742271977, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.456247329711914, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8852134943008423, "num_tokens": 811424479.0, "step": 21265 }, { "epoch": 2.7052537845057882, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.457551956176758, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8613787889480591, "num_tokens": 811464683.0, "step": 21266 }, { "epoch": 2.7053809947843783, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.395100593566895, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8897695541381836, "num_tokens": 811504801.0, "step": 21267 }, { "epoch": 2.7055082050629693, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.459426879882812, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8674054741859436, "num_tokens": 811546278.0, "step": 21268 }, { "epoch": 2.7056354153415594, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.395174026489258, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8611403703689575, "num_tokens": 811587019.0, "step": 21269 }, { "epoch": 2.7057626256201504, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.490182876586914, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8503393530845642, "num_tokens": 811619843.0, "step": 21270 }, { "epoch": 2.7058898358987404, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.406085014343262, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8802909255027771, "num_tokens": 811655797.0, "step": 21271 }, { "epoch": 2.706017046177331, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.462641716003418, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8665865063667297, "num_tokens": 811691626.0, "step": 21272 }, { "epoch": 2.7061442564559215, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.480154991149902, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8769205212593079, "num_tokens": 811726046.0, "step": 21273 }, { "epoch": 2.706271466734512, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.398170471191406, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8648902773857117, "num_tokens": 811765686.0, "step": 21274 }, { "epoch": 2.7063986770131025, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.463175773620605, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8825215101242065, "num_tokens": 811801143.0, "step": 21275 }, { "epoch": 2.706525887291693, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.4151611328125, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.880828857421875, "num_tokens": 811834807.0, "step": 21276 }, { "epoch": 2.7066530975702836, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.403297424316406, "learning_rate": 1e-06, "loss": 0.3435, "mean_token_accuracy": 0.8934733867645264, "num_tokens": 811865676.0, "step": 21277 }, { "epoch": 2.706780307848874, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.374246597290039, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8730324506759644, "num_tokens": 811901259.0, "step": 21278 }, { "epoch": 2.7069075181274647, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.398059844970703, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8614547252655029, "num_tokens": 811937768.0, "step": 21279 }, { "epoch": 2.707034728406055, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.412423133850098, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8744910955429077, "num_tokens": 811967805.0, "step": 21280 }, { "epoch": 2.7071619386846457, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.458227157592773, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8705602288246155, "num_tokens": 812007468.0, "step": 21281 }, { "epoch": 2.7072891489632362, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.408923149108887, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8757501840591431, "num_tokens": 812048127.0, "step": 21282 }, { "epoch": 2.7074163592418268, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.440017700195312, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8755009770393372, "num_tokens": 812083904.0, "step": 21283 }, { "epoch": 2.7075435695204173, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.454154014587402, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8795680999755859, "num_tokens": 812119643.0, "step": 21284 }, { "epoch": 2.707670779799008, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.445343971252441, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8738260269165039, "num_tokens": 812155338.0, "step": 21285 }, { "epoch": 2.7077979900775984, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.361948013305664, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8753474354743958, "num_tokens": 812193075.0, "step": 21286 }, { "epoch": 2.707925200356189, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.45504093170166, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8831688761711121, "num_tokens": 812228163.0, "step": 21287 }, { "epoch": 2.7080524106347794, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.445236206054688, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8690038919448853, "num_tokens": 812268565.0, "step": 21288 }, { "epoch": 2.70817962091337, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.327289581298828, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8795880675315857, "num_tokens": 812307292.0, "step": 21289 }, { "epoch": 2.7083068311919605, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.467310905456543, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8845518231391907, "num_tokens": 812346180.0, "step": 21290 }, { "epoch": 2.708434041470551, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.40298843383789, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.872448205947876, "num_tokens": 812376033.0, "step": 21291 }, { "epoch": 2.708561251749141, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.446115493774414, "learning_rate": 1e-06, "loss": 0.346, "mean_token_accuracy": 0.8919963240623474, "num_tokens": 812407925.0, "step": 21292 }, { "epoch": 2.708688462027732, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.430572509765625, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8804003000259399, "num_tokens": 812452354.0, "step": 21293 }, { "epoch": 2.708815672306322, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.342742919921875, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8741335868835449, "num_tokens": 812490015.0, "step": 21294 }, { "epoch": 2.708942882584913, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.400487899780273, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8708604574203491, "num_tokens": 812529919.0, "step": 21295 }, { "epoch": 2.709070092863503, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.417071342468262, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8869571089744568, "num_tokens": 812568541.0, "step": 21296 }, { "epoch": 2.7091973031420937, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.333429336547852, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8663370609283447, "num_tokens": 812612098.0, "step": 21297 }, { "epoch": 2.7093245134206843, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.457819938659668, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8713882565498352, "num_tokens": 812648843.0, "step": 21298 }, { "epoch": 2.709451723699275, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.291563987731934, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8754754066467285, "num_tokens": 812692080.0, "step": 21299 }, { "epoch": 2.7095789339778653, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.387267112731934, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8879685401916504, "num_tokens": 812729441.0, "step": 21300 }, { "epoch": 2.709706144256456, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.395397186279297, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8687856197357178, "num_tokens": 812768908.0, "step": 21301 }, { "epoch": 2.7098333545350464, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.366957664489746, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.874713659286499, "num_tokens": 812808384.0, "step": 21302 }, { "epoch": 2.709960564813637, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.343173027038574, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8664438724517822, "num_tokens": 812847658.0, "step": 21303 }, { "epoch": 2.7100877750922274, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.399048805236816, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8758280277252197, "num_tokens": 812889394.0, "step": 21304 }, { "epoch": 2.710214985370818, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.344401359558105, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8628419041633606, "num_tokens": 812924202.0, "step": 21305 }, { "epoch": 2.7103421956494085, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45151138305664, "learning_rate": 1e-06, "loss": 0.3345, "mean_token_accuracy": 0.8943723440170288, "num_tokens": 812954391.0, "step": 21306 }, { "epoch": 2.710469405927999, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.399486541748047, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8663657903671265, "num_tokens": 812989498.0, "step": 21307 }, { "epoch": 2.7105966162065895, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.356249809265137, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8815052509307861, "num_tokens": 813030968.0, "step": 21308 }, { "epoch": 2.71072382648518, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.352953910827637, "learning_rate": 1e-06, "loss": 0.3288, "mean_token_accuracy": 0.8965718746185303, "num_tokens": 813065843.0, "step": 21309 }, { "epoch": 2.7108510367637706, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.467164993286133, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8811059594154358, "num_tokens": 813107410.0, "step": 21310 }, { "epoch": 2.710978247042361, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.357646942138672, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.880585789680481, "num_tokens": 813155959.0, "step": 21311 }, { "epoch": 2.7111054573209517, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.30329418182373, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8814151287078857, "num_tokens": 813194133.0, "step": 21312 }, { "epoch": 2.711232667599542, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.466148376464844, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8622139692306519, "num_tokens": 813231015.0, "step": 21313 }, { "epoch": 2.7113598778781327, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.34184455871582, "learning_rate": 1e-06, "loss": 0.3424, "mean_token_accuracy": 0.8901325464248657, "num_tokens": 813271497.0, "step": 21314 }, { "epoch": 2.711487088156723, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.415273666381836, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.87666916847229, "num_tokens": 813305758.0, "step": 21315 }, { "epoch": 2.7116142984353138, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.379093170166016, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8733664155006409, "num_tokens": 813345958.0, "step": 21316 }, { "epoch": 2.711741508713904, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.265962600708008, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.877806544303894, "num_tokens": 813381630.0, "step": 21317 }, { "epoch": 2.711868718992495, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.381246566772461, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8757566809654236, "num_tokens": 813421320.0, "step": 21318 }, { "epoch": 2.711995929271085, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.34487247467041, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.868911862373352, "num_tokens": 813458323.0, "step": 21319 }, { "epoch": 2.712123139549676, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.376413345336914, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8572361469268799, "num_tokens": 813501436.0, "step": 21320 }, { "epoch": 2.712250349828266, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.336965560913086, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.874960720539093, "num_tokens": 813545942.0, "step": 21321 }, { "epoch": 2.7123775601068565, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.365443229675293, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.870302677154541, "num_tokens": 813582202.0, "step": 21322 }, { "epoch": 2.712504770385447, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.465119361877441, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.860093891620636, "num_tokens": 813621711.0, "step": 21323 }, { "epoch": 2.7126319806640375, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.300613403320312, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8697468638420105, "num_tokens": 813665404.0, "step": 21324 }, { "epoch": 2.712759190942628, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.450721740722656, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8822337985038757, "num_tokens": 813697470.0, "step": 21325 }, { "epoch": 2.7128864012212186, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.341097831726074, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8819782137870789, "num_tokens": 813733308.0, "step": 21326 }, { "epoch": 2.713013611499809, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.434867858886719, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8623026013374329, "num_tokens": 813770881.0, "step": 21327 }, { "epoch": 2.7131408217783997, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45893669128418, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8804795742034912, "num_tokens": 813806495.0, "step": 21328 }, { "epoch": 2.71326803205699, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.360649108886719, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8799739480018616, "num_tokens": 813842408.0, "step": 21329 }, { "epoch": 2.7133952423355807, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.353788375854492, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8711416721343994, "num_tokens": 813883041.0, "step": 21330 }, { "epoch": 2.7135224526141712, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.456748962402344, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8709149956703186, "num_tokens": 813922901.0, "step": 21331 }, { "epoch": 2.7136496628927618, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.367761611938477, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.876736044883728, "num_tokens": 813961948.0, "step": 21332 }, { "epoch": 2.7137768731713523, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.489012718200684, "learning_rate": 1e-06, "loss": 0.3545, "mean_token_accuracy": 0.8891993165016174, "num_tokens": 813999194.0, "step": 21333 }, { "epoch": 2.713904083449943, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.35663890838623, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8643496632575989, "num_tokens": 814038986.0, "step": 21334 }, { "epoch": 2.7140312937285334, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.416716575622559, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8885958194732666, "num_tokens": 814074351.0, "step": 21335 }, { "epoch": 2.714158504007124, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.455574989318848, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8751084804534912, "num_tokens": 814109255.0, "step": 21336 }, { "epoch": 2.7142857142857144, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.461220741271973, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.871039867401123, "num_tokens": 814146601.0, "step": 21337 }, { "epoch": 2.714412924564305, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.459157943725586, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8795841336250305, "num_tokens": 814181299.0, "step": 21338 }, { "epoch": 2.7145401348428955, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.410667419433594, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.88301682472229, "num_tokens": 814218178.0, "step": 21339 }, { "epoch": 2.7146673451214856, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.441410064697266, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8771013617515564, "num_tokens": 814254161.0, "step": 21340 }, { "epoch": 2.7147945554000765, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.41274642944336, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8760782480239868, "num_tokens": 814290164.0, "step": 21341 }, { "epoch": 2.7149217656786666, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.391077995300293, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8549482226371765, "num_tokens": 814334459.0, "step": 21342 }, { "epoch": 2.7150489759572576, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.418208122253418, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8677042722702026, "num_tokens": 814374291.0, "step": 21343 }, { "epoch": 2.7151761862358477, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.4010009765625, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8697988986968994, "num_tokens": 814413196.0, "step": 21344 }, { "epoch": 2.7153033965144386, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.372549057006836, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.873767614364624, "num_tokens": 814455312.0, "step": 21345 }, { "epoch": 2.7154306067930287, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.363442420959473, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8711703419685364, "num_tokens": 814496586.0, "step": 21346 }, { "epoch": 2.7155578170716193, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.439168930053711, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8802022933959961, "num_tokens": 814531566.0, "step": 21347 }, { "epoch": 2.71568502735021, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.352086067199707, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8815401792526245, "num_tokens": 814566168.0, "step": 21348 }, { "epoch": 2.7158122376288003, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.404263496398926, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8747140169143677, "num_tokens": 814604927.0, "step": 21349 }, { "epoch": 2.715939447907391, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.359968185424805, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8887884616851807, "num_tokens": 814641401.0, "step": 21350 }, { "epoch": 2.7160666581859814, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.377586364746094, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8808320164680481, "num_tokens": 814674252.0, "step": 21351 }, { "epoch": 2.716193868464572, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.419172286987305, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8743207454681396, "num_tokens": 814704965.0, "step": 21352 }, { "epoch": 2.7163210787431624, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.433257102966309, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8473857641220093, "num_tokens": 814743740.0, "step": 21353 }, { "epoch": 2.716448289021753, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.381983757019043, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.881756067276001, "num_tokens": 814784598.0, "step": 21354 }, { "epoch": 2.7165754993003435, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.418435096740723, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8685837984085083, "num_tokens": 814821437.0, "step": 21355 }, { "epoch": 2.716702709578934, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.448697090148926, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8748260736465454, "num_tokens": 814861638.0, "step": 21356 }, { "epoch": 2.7168299198575245, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.379617691040039, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8763456344604492, "num_tokens": 814893799.0, "step": 21357 }, { "epoch": 2.716957130136115, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.41832160949707, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.882354199886322, "num_tokens": 814930726.0, "step": 21358 }, { "epoch": 2.7170843404147056, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.345013618469238, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8737301826477051, "num_tokens": 814970041.0, "step": 21359 }, { "epoch": 2.717211550693296, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.450418472290039, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8785220384597778, "num_tokens": 815008587.0, "step": 21360 }, { "epoch": 2.7173387609718866, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.429184913635254, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8630433082580566, "num_tokens": 815047659.0, "step": 21361 }, { "epoch": 2.717465971250477, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.421006202697754, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8753572106361389, "num_tokens": 815082631.0, "step": 21362 }, { "epoch": 2.7175931815290677, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.463167190551758, "learning_rate": 1e-06, "loss": 0.3381, "mean_token_accuracy": 0.8928782939910889, "num_tokens": 815117568.0, "step": 21363 }, { "epoch": 2.7177203918076582, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.389649391174316, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8754328489303589, "num_tokens": 815155019.0, "step": 21364 }, { "epoch": 2.7178476020862483, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.450257301330566, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8750499486923218, "num_tokens": 815195386.0, "step": 21365 }, { "epoch": 2.7179748123648393, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.365558624267578, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.879463791847229, "num_tokens": 815228304.0, "step": 21366 }, { "epoch": 2.7181020226434294, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.37547492980957, "learning_rate": 1e-06, "loss": 0.3315, "mean_token_accuracy": 0.8948070406913757, "num_tokens": 815260111.0, "step": 21367 }, { "epoch": 2.7182292329220203, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.350974082946777, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8742154836654663, "num_tokens": 815297272.0, "step": 21368 }, { "epoch": 2.7183564432006104, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.377898216247559, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8735620379447937, "num_tokens": 815336016.0, "step": 21369 }, { "epoch": 2.718483653479201, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.48226547241211, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8490349054336548, "num_tokens": 815373940.0, "step": 21370 }, { "epoch": 2.7186108637577915, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.36672306060791, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8647091388702393, "num_tokens": 815410836.0, "step": 21371 }, { "epoch": 2.718738074036382, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.420088768005371, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8807783126831055, "num_tokens": 815449577.0, "step": 21372 }, { "epoch": 2.7188652843149725, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.41555118560791, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8782725930213928, "num_tokens": 815485869.0, "step": 21373 }, { "epoch": 2.718992494593563, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.44393539428711, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.85670405626297, "num_tokens": 815523614.0, "step": 21374 }, { "epoch": 2.7191197048721536, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.410490036010742, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8550205826759338, "num_tokens": 815561835.0, "step": 21375 }, { "epoch": 2.719246915150744, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.385273933410645, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8603127002716064, "num_tokens": 815605596.0, "step": 21376 }, { "epoch": 2.7193741254293347, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.430044174194336, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8694620132446289, "num_tokens": 815638720.0, "step": 21377 }, { "epoch": 2.719501335707925, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.402137756347656, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8892942667007446, "num_tokens": 815673996.0, "step": 21378 }, { "epoch": 2.7196285459865157, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.40811538696289, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8628115653991699, "num_tokens": 815716571.0, "step": 21379 }, { "epoch": 2.7197557562651062, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.448301315307617, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.882155179977417, "num_tokens": 815748973.0, "step": 21380 }, { "epoch": 2.7198829665436968, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.380695343017578, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8768584132194519, "num_tokens": 815794754.0, "step": 21381 }, { "epoch": 2.7200101768222873, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.470891952514648, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8782253265380859, "num_tokens": 815831051.0, "step": 21382 }, { "epoch": 2.720137387100878, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.377760887145996, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8765946626663208, "num_tokens": 815874291.0, "step": 21383 }, { "epoch": 2.7202645973794684, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.452347755432129, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8724644780158997, "num_tokens": 815912484.0, "step": 21384 }, { "epoch": 2.720391807658059, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.465914726257324, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8592345118522644, "num_tokens": 815948626.0, "step": 21385 }, { "epoch": 2.7205190179366494, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.389439582824707, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8836607933044434, "num_tokens": 815983154.0, "step": 21386 }, { "epoch": 2.72064622821524, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.461179733276367, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8774884939193726, "num_tokens": 816017524.0, "step": 21387 }, { "epoch": 2.7207734384938305, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.465733528137207, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8696338534355164, "num_tokens": 816055204.0, "step": 21388 }, { "epoch": 2.720900648772421, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.438738822937012, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8823398351669312, "num_tokens": 816086671.0, "step": 21389 }, { "epoch": 2.721027859051011, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.465929985046387, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8561856150627136, "num_tokens": 816114173.0, "step": 21390 }, { "epoch": 2.721155069329602, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.439358711242676, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8869839906692505, "num_tokens": 816156434.0, "step": 21391 }, { "epoch": 2.721282279608192, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.442915916442871, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.870693564414978, "num_tokens": 816191037.0, "step": 21392 }, { "epoch": 2.721409489886783, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.511214256286621, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.888105571269989, "num_tokens": 816227031.0, "step": 21393 }, { "epoch": 2.721536700165373, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.41454029083252, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8781218528747559, "num_tokens": 816264308.0, "step": 21394 }, { "epoch": 2.7216639104439637, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.424702644348145, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8730637431144714, "num_tokens": 816300955.0, "step": 21395 }, { "epoch": 2.7217911207225542, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.463494300842285, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.876049280166626, "num_tokens": 816332881.0, "step": 21396 }, { "epoch": 2.7219183310011448, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.3334321975708, "learning_rate": 1e-06, "loss": 0.3406, "mean_token_accuracy": 0.8933861255645752, "num_tokens": 816367446.0, "step": 21397 }, { "epoch": 2.7220455412797353, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.446722984313965, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8589916825294495, "num_tokens": 816407286.0, "step": 21398 }, { "epoch": 2.722172751558326, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.439661026000977, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.883660078048706, "num_tokens": 816445277.0, "step": 21399 }, { "epoch": 2.7222999618369164, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.455144882202148, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8803561329841614, "num_tokens": 816481638.0, "step": 21400 }, { "epoch": 2.722427172115507, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.43712043762207, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8762983083724976, "num_tokens": 816521386.0, "step": 21401 }, { "epoch": 2.7225543823940974, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.385208129882812, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.885986328125, "num_tokens": 816554760.0, "step": 21402 }, { "epoch": 2.722681592672688, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.470417976379395, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8831043243408203, "num_tokens": 816595195.0, "step": 21403 }, { "epoch": 2.7228088029512785, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.40679931640625, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8765741586685181, "num_tokens": 816633046.0, "step": 21404 }, { "epoch": 2.722936013229869, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.411250114440918, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8889522552490234, "num_tokens": 816675283.0, "step": 21405 }, { "epoch": 2.7230632235084595, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.43826961517334, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8660457730293274, "num_tokens": 816716499.0, "step": 21406 }, { "epoch": 2.72319043378705, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.457043647766113, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.869492769241333, "num_tokens": 816761825.0, "step": 21407 }, { "epoch": 2.7233176440656406, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.4093656539917, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8815112113952637, "num_tokens": 816792080.0, "step": 21408 }, { "epoch": 2.723444854344231, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.405993461608887, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8657382726669312, "num_tokens": 816832121.0, "step": 21409 }, { "epoch": 2.7235720646228216, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.441596031188965, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8683273792266846, "num_tokens": 816874032.0, "step": 21410 }, { "epoch": 2.723699274901412, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.4505615234375, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8570929765701294, "num_tokens": 816920471.0, "step": 21411 }, { "epoch": 2.7238264851800027, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.45580768585205, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8761783838272095, "num_tokens": 816958042.0, "step": 21412 }, { "epoch": 2.723953695458593, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45999813079834, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8667840361595154, "num_tokens": 816992158.0, "step": 21413 }, { "epoch": 2.7240809057371838, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.509079933166504, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8642233610153198, "num_tokens": 817032552.0, "step": 21414 }, { "epoch": 2.724208116015774, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.335185050964355, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8725521564483643, "num_tokens": 817072240.0, "step": 21415 }, { "epoch": 2.724335326294365, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.515335083007812, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8653621077537537, "num_tokens": 817104741.0, "step": 21416 }, { "epoch": 2.724462536572955, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.395782470703125, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8628966808319092, "num_tokens": 817143876.0, "step": 21417 }, { "epoch": 2.724589746851546, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.401396751403809, "learning_rate": 1e-06, "loss": 0.3227, "mean_token_accuracy": 0.8974561095237732, "num_tokens": 817183647.0, "step": 21418 }, { "epoch": 2.724716957130136, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.5479736328125, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8698188066482544, "num_tokens": 817214520.0, "step": 21419 }, { "epoch": 2.7248441674087265, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.444961547851562, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8658769130706787, "num_tokens": 817256936.0, "step": 21420 }, { "epoch": 2.724971377687317, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.48967170715332, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8750375509262085, "num_tokens": 817292840.0, "step": 21421 }, { "epoch": 2.7250985879659075, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.40201473236084, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8710595965385437, "num_tokens": 817332872.0, "step": 21422 }, { "epoch": 2.725225798244498, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.42851448059082, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8827229738235474, "num_tokens": 817367668.0, "step": 21423 }, { "epoch": 2.7253530085230886, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.410985946655273, "learning_rate": 1e-06, "loss": 0.3418, "mean_token_accuracy": 0.8929907083511353, "num_tokens": 817407726.0, "step": 21424 }, { "epoch": 2.725480218801679, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.496082305908203, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8853899836540222, "num_tokens": 817441003.0, "step": 21425 }, { "epoch": 2.7256074290802697, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.416242599487305, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8801488876342773, "num_tokens": 817482587.0, "step": 21426 }, { "epoch": 2.72573463935886, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.478293418884277, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8796103000640869, "num_tokens": 817519727.0, "step": 21427 }, { "epoch": 2.7258618496374507, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45321273803711, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8721826672554016, "num_tokens": 817553809.0, "step": 21428 }, { "epoch": 2.7259890599160412, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.462045669555664, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.859098494052887, "num_tokens": 817588893.0, "step": 21429 }, { "epoch": 2.7261162701946318, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.464688301086426, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8816899061203003, "num_tokens": 817627789.0, "step": 21430 }, { "epoch": 2.7262434804732223, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.409337997436523, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8655929565429688, "num_tokens": 817670048.0, "step": 21431 }, { "epoch": 2.726370690751813, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.490239143371582, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8806101083755493, "num_tokens": 817707108.0, "step": 21432 }, { "epoch": 2.7264979010304033, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.423277854919434, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8735448122024536, "num_tokens": 817748126.0, "step": 21433 }, { "epoch": 2.726625111308994, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.422441482543945, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8780702352523804, "num_tokens": 817781578.0, "step": 21434 }, { "epoch": 2.7267523215875844, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.427398681640625, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8823548555374146, "num_tokens": 817822143.0, "step": 21435 }, { "epoch": 2.726879531866175, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.478394508361816, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.885664701461792, "num_tokens": 817858456.0, "step": 21436 }, { "epoch": 2.7270067421447655, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.420409202575684, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8819867372512817, "num_tokens": 817892863.0, "step": 21437 }, { "epoch": 2.7271339524233555, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.409372329711914, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8795740604400635, "num_tokens": 817931120.0, "step": 21438 }, { "epoch": 2.7272611627019465, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.511752128601074, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8788182139396667, "num_tokens": 817963076.0, "step": 21439 }, { "epoch": 2.7273883729805366, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.459539413452148, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.876519501209259, "num_tokens": 818003844.0, "step": 21440 }, { "epoch": 2.7275155832591276, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.482476234436035, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.877976655960083, "num_tokens": 818048533.0, "step": 21441 }, { "epoch": 2.7276427935377177, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.462298393249512, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.871324896812439, "num_tokens": 818083290.0, "step": 21442 }, { "epoch": 2.7277700038163086, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.456451416015625, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8828809857368469, "num_tokens": 818120037.0, "step": 21443 }, { "epoch": 2.7278972140948987, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45814323425293, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8758721351623535, "num_tokens": 818164122.0, "step": 21444 }, { "epoch": 2.7280244243734892, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.454174041748047, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8768258094787598, "num_tokens": 818198065.0, "step": 21445 }, { "epoch": 2.7281516346520798, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.415987014770508, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8774068355560303, "num_tokens": 818236843.0, "step": 21446 }, { "epoch": 2.7282788449306703, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.43265438079834, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8761203289031982, "num_tokens": 818271300.0, "step": 21447 }, { "epoch": 2.728406055209261, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.423999786376953, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8614431023597717, "num_tokens": 818311799.0, "step": 21448 }, { "epoch": 2.7285332654878514, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.441397666931152, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.880731463432312, "num_tokens": 818346662.0, "step": 21449 }, { "epoch": 2.728660475766442, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.417072296142578, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8794165849685669, "num_tokens": 818384465.0, "step": 21450 }, { "epoch": 2.7287876860450324, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.478175163269043, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8727982044219971, "num_tokens": 818425625.0, "step": 21451 }, { "epoch": 2.728914896323623, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.437079429626465, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8799622058868408, "num_tokens": 818458544.0, "step": 21452 }, { "epoch": 2.7290421066022135, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.434196472167969, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8708670139312744, "num_tokens": 818503507.0, "step": 21453 }, { "epoch": 2.729169316880804, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.3880615234375, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8810751438140869, "num_tokens": 818545411.0, "step": 21454 }, { "epoch": 2.7292965271593945, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.497332572937012, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8709856271743774, "num_tokens": 818582100.0, "step": 21455 }, { "epoch": 2.729423737437985, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.379623413085938, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.872079074382782, "num_tokens": 818615549.0, "step": 21456 }, { "epoch": 2.7295509477165756, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.402400016784668, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8794417381286621, "num_tokens": 818652476.0, "step": 21457 }, { "epoch": 2.729678157995166, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.498727798461914, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8793734908103943, "num_tokens": 818692435.0, "step": 21458 }, { "epoch": 2.7298053682737566, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.368581771850586, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8739139437675476, "num_tokens": 818733206.0, "step": 21459 }, { "epoch": 2.729932578552347, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.42170238494873, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8792247772216797, "num_tokens": 818771717.0, "step": 21460 }, { "epoch": 2.7300597888309377, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.444795608520508, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8758015036582947, "num_tokens": 818806999.0, "step": 21461 }, { "epoch": 2.7301869991095282, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.394355773925781, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8799701929092407, "num_tokens": 818846226.0, "step": 21462 }, { "epoch": 2.7303142093881183, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.39319896697998, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8814305663108826, "num_tokens": 818885667.0, "step": 21463 }, { "epoch": 2.7304414196667093, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.405438423156738, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8808099627494812, "num_tokens": 818924009.0, "step": 21464 }, { "epoch": 2.7305686299452994, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.361251831054688, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8843529224395752, "num_tokens": 818962614.0, "step": 21465 }, { "epoch": 2.7306958402238903, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.380914688110352, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.87764573097229, "num_tokens": 819004486.0, "step": 21466 }, { "epoch": 2.7308230505024804, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.405722618103027, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8696830868721008, "num_tokens": 819048364.0, "step": 21467 }, { "epoch": 2.730950260781071, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.37692928314209, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8724989891052246, "num_tokens": 819086250.0, "step": 21468 }, { "epoch": 2.7310774710596615, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.497182846069336, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8864489793777466, "num_tokens": 819126705.0, "step": 21469 }, { "epoch": 2.731204681338252, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.409737586975098, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8769559860229492, "num_tokens": 819165967.0, "step": 21470 }, { "epoch": 2.7313318916168425, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.378447532653809, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8718419075012207, "num_tokens": 819208066.0, "step": 21471 }, { "epoch": 2.731459101895433, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.482893943786621, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8728739619255066, "num_tokens": 819246269.0, "step": 21472 }, { "epoch": 2.7315863121740236, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.386116981506348, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8765532970428467, "num_tokens": 819286027.0, "step": 21473 }, { "epoch": 2.731713522452614, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.382218360900879, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8842782974243164, "num_tokens": 819319994.0, "step": 21474 }, { "epoch": 2.7318407327312046, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.461633682250977, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8716365098953247, "num_tokens": 819360113.0, "step": 21475 }, { "epoch": 2.731967943009795, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.354511260986328, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8853872418403625, "num_tokens": 819394575.0, "step": 21476 }, { "epoch": 2.7320951532883857, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.3951416015625, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8705911636352539, "num_tokens": 819434545.0, "step": 21477 }, { "epoch": 2.7322223635669762, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.422492027282715, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8855164051055908, "num_tokens": 819470699.0, "step": 21478 }, { "epoch": 2.7323495738455668, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.403757095336914, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8702331781387329, "num_tokens": 819514447.0, "step": 21479 }, { "epoch": 2.7324767841241573, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.41343879699707, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8700073957443237, "num_tokens": 819553434.0, "step": 21480 }, { "epoch": 2.732603994402748, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.403731346130371, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8686323761940002, "num_tokens": 819591594.0, "step": 21481 }, { "epoch": 2.7327312046813383, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.355843544006348, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8716561198234558, "num_tokens": 819634530.0, "step": 21482 }, { "epoch": 2.732858414959929, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.398550987243652, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8664923906326294, "num_tokens": 819669358.0, "step": 21483 }, { "epoch": 2.7329856252385194, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.3691987991333, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8749207258224487, "num_tokens": 819711336.0, "step": 21484 }, { "epoch": 2.73311283551711, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.422164916992188, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8665044903755188, "num_tokens": 819746973.0, "step": 21485 }, { "epoch": 2.7332400457957005, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.323493957519531, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8851087689399719, "num_tokens": 819780305.0, "step": 21486 }, { "epoch": 2.733367256074291, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.452019691467285, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8835646510124207, "num_tokens": 819817326.0, "step": 21487 }, { "epoch": 2.733494466352881, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.357301712036133, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8699599504470825, "num_tokens": 819857597.0, "step": 21488 }, { "epoch": 2.733621676631472, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.437424659729004, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8783769011497498, "num_tokens": 819893471.0, "step": 21489 }, { "epoch": 2.733748886910062, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.449459075927734, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8580408096313477, "num_tokens": 819934175.0, "step": 21490 }, { "epoch": 2.733876097188653, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.41097354888916, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8839301466941833, "num_tokens": 819971451.0, "step": 21491 }, { "epoch": 2.734003307467243, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.398813247680664, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8697543740272522, "num_tokens": 820011566.0, "step": 21492 }, { "epoch": 2.7341305177458337, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.473724365234375, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8523651361465454, "num_tokens": 820047944.0, "step": 21493 }, { "epoch": 2.7342577280244242, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.353551864624023, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8639213442802429, "num_tokens": 820083416.0, "step": 21494 }, { "epoch": 2.7343849383030148, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.449544906616211, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8673630356788635, "num_tokens": 820122476.0, "step": 21495 }, { "epoch": 2.7345121485816053, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.33878231048584, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8808565139770508, "num_tokens": 820159866.0, "step": 21496 }, { "epoch": 2.734639358860196, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.384807586669922, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8752833008766174, "num_tokens": 820202710.0, "step": 21497 }, { "epoch": 2.7347665691387864, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.416497230529785, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8733187913894653, "num_tokens": 820239156.0, "step": 21498 }, { "epoch": 2.734893779417377, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.461009979248047, "learning_rate": 1e-06, "loss": 0.333, "mean_token_accuracy": 0.8929300308227539, "num_tokens": 820274010.0, "step": 21499 }, { "epoch": 2.7350209896959674, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.394582748413086, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8730903267860413, "num_tokens": 820319171.0, "step": 21500 }, { "epoch": 2.735148199974558, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.471240043640137, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8610249161720276, "num_tokens": 820355038.0, "step": 21501 }, { "epoch": 2.7352754102531485, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.48969554901123, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8716681003570557, "num_tokens": 820390973.0, "step": 21502 }, { "epoch": 2.735402620531739, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.417055130004883, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8701625466346741, "num_tokens": 820424543.0, "step": 21503 }, { "epoch": 2.7355298308103295, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.458602905273438, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8732954859733582, "num_tokens": 820462702.0, "step": 21504 }, { "epoch": 2.73565704108892, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.387638092041016, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8844218850135803, "num_tokens": 820498840.0, "step": 21505 }, { "epoch": 2.7357842513675106, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.467694282531738, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8682501316070557, "num_tokens": 820534083.0, "step": 21506 }, { "epoch": 2.735911461646101, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.452482223510742, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8789891600608826, "num_tokens": 820572479.0, "step": 21507 }, { "epoch": 2.7360386719246916, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.385618209838867, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8632651567459106, "num_tokens": 820610283.0, "step": 21508 }, { "epoch": 2.736165882203282, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.419100761413574, "learning_rate": 1e-06, "loss": 0.3509, "mean_token_accuracy": 0.8908727169036865, "num_tokens": 820647580.0, "step": 21509 }, { "epoch": 2.7362930924818727, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.437302589416504, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8808411955833435, "num_tokens": 820686774.0, "step": 21510 }, { "epoch": 2.7364203027604628, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.374306678771973, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.884291410446167, "num_tokens": 820724635.0, "step": 21511 }, { "epoch": 2.7365475130390537, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.514219284057617, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8816030025482178, "num_tokens": 820761467.0, "step": 21512 }, { "epoch": 2.736674723317644, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.435811996459961, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8623371124267578, "num_tokens": 820801001.0, "step": 21513 }, { "epoch": 2.736801933596235, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.37907886505127, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8697646260261536, "num_tokens": 820841028.0, "step": 21514 }, { "epoch": 2.736929143874825, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.498025894165039, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8827446103096008, "num_tokens": 820881613.0, "step": 21515 }, { "epoch": 2.737056354153416, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.3672456741333, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8829464912414551, "num_tokens": 820918205.0, "step": 21516 }, { "epoch": 2.737183564432006, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.423755645751953, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8679416179656982, "num_tokens": 820950751.0, "step": 21517 }, { "epoch": 2.7373107747105965, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.385619163513184, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8769949674606323, "num_tokens": 820987385.0, "step": 21518 }, { "epoch": 2.737437984989187, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.485054969787598, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8682281970977783, "num_tokens": 821026273.0, "step": 21519 }, { "epoch": 2.7375651952677775, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.475833892822266, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8674197196960449, "num_tokens": 821067978.0, "step": 21520 }, { "epoch": 2.737692405546368, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.408766746520996, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8804092407226562, "num_tokens": 821111925.0, "step": 21521 }, { "epoch": 2.7378196158249586, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.457600593566895, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8788004517555237, "num_tokens": 821151019.0, "step": 21522 }, { "epoch": 2.737946826103549, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.487486839294434, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8822799921035767, "num_tokens": 821185520.0, "step": 21523 }, { "epoch": 2.7380740363821396, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.430685043334961, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.877110481262207, "num_tokens": 821224273.0, "step": 21524 }, { "epoch": 2.73820124666073, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.420109748840332, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8809703588485718, "num_tokens": 821263541.0, "step": 21525 }, { "epoch": 2.7383284569393207, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.424509048461914, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8637772798538208, "num_tokens": 821304053.0, "step": 21526 }, { "epoch": 2.7384556672179112, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.405655860900879, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8762230277061462, "num_tokens": 821339710.0, "step": 21527 }, { "epoch": 2.7385828774965018, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.339835166931152, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.875619113445282, "num_tokens": 821378377.0, "step": 21528 }, { "epoch": 2.7387100877750923, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.463675498962402, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8837328553199768, "num_tokens": 821417434.0, "step": 21529 }, { "epoch": 2.738837298053683, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.31749153137207, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8718934655189514, "num_tokens": 821464323.0, "step": 21530 }, { "epoch": 2.7389645083322733, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.507501602172852, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8728264570236206, "num_tokens": 821509891.0, "step": 21531 }, { "epoch": 2.739091718610864, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.461990356445312, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8645473122596741, "num_tokens": 821557309.0, "step": 21532 }, { "epoch": 2.7392189288894544, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.489893913269043, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8702365756034851, "num_tokens": 821597751.0, "step": 21533 }, { "epoch": 2.739346139168045, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.387250900268555, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8864312767982483, "num_tokens": 821637961.0, "step": 21534 }, { "epoch": 2.7394733494466355, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.45406723022461, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8583760261535645, "num_tokens": 821684728.0, "step": 21535 }, { "epoch": 2.7396005597252255, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.472259521484375, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8797397613525391, "num_tokens": 821715359.0, "step": 21536 }, { "epoch": 2.7397277700038165, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.462164878845215, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8689957857131958, "num_tokens": 821755540.0, "step": 21537 }, { "epoch": 2.7398549802824066, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.333662033081055, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8764798641204834, "num_tokens": 821788506.0, "step": 21538 }, { "epoch": 2.7399821905609976, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.418539047241211, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8748138546943665, "num_tokens": 821826555.0, "step": 21539 }, { "epoch": 2.7401094008395877, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.3744535446167, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8651798367500305, "num_tokens": 821866160.0, "step": 21540 }, { "epoch": 2.7402366111181786, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.488167762756348, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8803737163543701, "num_tokens": 821905595.0, "step": 21541 }, { "epoch": 2.7403638213967687, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.42977237701416, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8665611147880554, "num_tokens": 821944769.0, "step": 21542 }, { "epoch": 2.7404910316753592, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.39101791381836, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8827334642410278, "num_tokens": 821988889.0, "step": 21543 }, { "epoch": 2.7406182419539498, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.445208549499512, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8784183263778687, "num_tokens": 822025343.0, "step": 21544 }, { "epoch": 2.7407454522325403, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.386067390441895, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.863446831703186, "num_tokens": 822067651.0, "step": 21545 }, { "epoch": 2.740872662511131, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.466144561767578, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8648305535316467, "num_tokens": 822102885.0, "step": 21546 }, { "epoch": 2.7409998727897213, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.331870079040527, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8667522072792053, "num_tokens": 822142515.0, "step": 21547 }, { "epoch": 2.741127083068312, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.409786224365234, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8712641000747681, "num_tokens": 822180745.0, "step": 21548 }, { "epoch": 2.7412542933469024, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.393492698669434, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8814309239387512, "num_tokens": 822215117.0, "step": 21549 }, { "epoch": 2.741381503625493, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.445618629455566, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8646821975708008, "num_tokens": 822258883.0, "step": 21550 }, { "epoch": 2.7415087139040835, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.39500904083252, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8735207915306091, "num_tokens": 822295254.0, "step": 21551 }, { "epoch": 2.741635924182674, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.36075210571289, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8767440319061279, "num_tokens": 822335769.0, "step": 21552 }, { "epoch": 2.7417631344612645, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.362338066101074, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8837101459503174, "num_tokens": 822376873.0, "step": 21553 }, { "epoch": 2.741890344739855, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.406713485717773, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8722346425056458, "num_tokens": 822419868.0, "step": 21554 }, { "epoch": 2.7420175550184456, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.320830345153809, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8736895322799683, "num_tokens": 822456757.0, "step": 21555 }, { "epoch": 2.742144765297036, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.40878677368164, "learning_rate": 1e-06, "loss": 0.3499, "mean_token_accuracy": 0.8877653479576111, "num_tokens": 822492772.0, "step": 21556 }, { "epoch": 2.7422719755756266, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.439313888549805, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8585402369499207, "num_tokens": 822535400.0, "step": 21557 }, { "epoch": 2.742399185854217, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.409122467041016, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8799586892127991, "num_tokens": 822568313.0, "step": 21558 }, { "epoch": 2.7425263961328077, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.369451522827148, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8814125061035156, "num_tokens": 822608487.0, "step": 21559 }, { "epoch": 2.742653606411398, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.478577613830566, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8749293088912964, "num_tokens": 822647938.0, "step": 21560 }, { "epoch": 2.7427808166899883, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.322647094726562, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.870608389377594, "num_tokens": 822685974.0, "step": 21561 }, { "epoch": 2.7429080269685793, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.468304634094238, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8768218159675598, "num_tokens": 822719166.0, "step": 21562 }, { "epoch": 2.7430352372471694, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.38208293914795, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8641311526298523, "num_tokens": 822758364.0, "step": 21563 }, { "epoch": 2.7431624475257603, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.39297103881836, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8823561668395996, "num_tokens": 822798002.0, "step": 21564 }, { "epoch": 2.7432896578043504, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.340048789978027, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8755196332931519, "num_tokens": 822837958.0, "step": 21565 }, { "epoch": 2.743416868082941, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.464176177978516, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8751220107078552, "num_tokens": 822873076.0, "step": 21566 }, { "epoch": 2.7435440783615315, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.448161125183105, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8737289309501648, "num_tokens": 822904693.0, "step": 21567 }, { "epoch": 2.743671288640122, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.42214298248291, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8695958852767944, "num_tokens": 822941192.0, "step": 21568 }, { "epoch": 2.7437984989187125, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.329277992248535, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8627163767814636, "num_tokens": 822982166.0, "step": 21569 }, { "epoch": 2.743925709197303, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.418684005737305, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8686388731002808, "num_tokens": 823021630.0, "step": 21570 }, { "epoch": 2.7440529194758936, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.422636985778809, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8744677305221558, "num_tokens": 823062133.0, "step": 21571 }, { "epoch": 2.744180129754484, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.35921573638916, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8718149065971375, "num_tokens": 823105163.0, "step": 21572 }, { "epoch": 2.7443073400330746, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.420310020446777, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8544490337371826, "num_tokens": 823146878.0, "step": 21573 }, { "epoch": 2.744434550311665, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.446288108825684, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.891819953918457, "num_tokens": 823182216.0, "step": 21574 }, { "epoch": 2.7445617605902557, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.420745849609375, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8818426132202148, "num_tokens": 823227838.0, "step": 21575 }, { "epoch": 2.7446889708688462, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.461747169494629, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8666050434112549, "num_tokens": 823268609.0, "step": 21576 }, { "epoch": 2.7448161811474368, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.409676551818848, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8831871151924133, "num_tokens": 823309426.0, "step": 21577 }, { "epoch": 2.7449433914260273, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.447141647338867, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8699623346328735, "num_tokens": 823346049.0, "step": 21578 }, { "epoch": 2.745070601704618, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.494950294494629, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8680498600006104, "num_tokens": 823378915.0, "step": 21579 }, { "epoch": 2.7451978119832083, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.399964332580566, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8640113472938538, "num_tokens": 823422985.0, "step": 21580 }, { "epoch": 2.745325022261799, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.492159843444824, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8618152141571045, "num_tokens": 823460703.0, "step": 21581 }, { "epoch": 2.7454522325403894, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.462200164794922, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8776167631149292, "num_tokens": 823499651.0, "step": 21582 }, { "epoch": 2.74557944281898, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.426664352416992, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8736013174057007, "num_tokens": 823537044.0, "step": 21583 }, { "epoch": 2.7457066530975704, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.471742630004883, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8872175216674805, "num_tokens": 823571832.0, "step": 21584 }, { "epoch": 2.745833863376161, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.396855354309082, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8808773159980774, "num_tokens": 823609852.0, "step": 21585 }, { "epoch": 2.745961073654751, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.502774238586426, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8782506585121155, "num_tokens": 823644419.0, "step": 21586 }, { "epoch": 2.746088283933342, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.386832237243652, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8797401189804077, "num_tokens": 823682451.0, "step": 21587 }, { "epoch": 2.746215494211932, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.410262107849121, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8850979804992676, "num_tokens": 823723051.0, "step": 21588 }, { "epoch": 2.746342704490523, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.499309539794922, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8716850876808167, "num_tokens": 823761353.0, "step": 21589 }, { "epoch": 2.746469914769113, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.412476539611816, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8597238063812256, "num_tokens": 823800281.0, "step": 21590 }, { "epoch": 2.7465971250477037, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.445913314819336, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8875038623809814, "num_tokens": 823843255.0, "step": 21591 }, { "epoch": 2.7467243353262942, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.419984817504883, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8716351389884949, "num_tokens": 823880736.0, "step": 21592 }, { "epoch": 2.7468515456048848, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.508221626281738, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8719800710678101, "num_tokens": 823923232.0, "step": 21593 }, { "epoch": 2.7469787558834753, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.391334533691406, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8775115013122559, "num_tokens": 823957488.0, "step": 21594 }, { "epoch": 2.747105966162066, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.421772956848145, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8697302341461182, "num_tokens": 823999568.0, "step": 21595 }, { "epoch": 2.7472331764406563, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.419204711914062, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8863637447357178, "num_tokens": 824036947.0, "step": 21596 }, { "epoch": 2.747360386719247, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.508889198303223, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8869507312774658, "num_tokens": 824067966.0, "step": 21597 }, { "epoch": 2.7474875969978374, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.40359115600586, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.872717022895813, "num_tokens": 824112216.0, "step": 21598 }, { "epoch": 2.747614807276428, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.463115692138672, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8844496011734009, "num_tokens": 824148749.0, "step": 21599 }, { "epoch": 2.7477420175550185, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.446925163269043, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8727765679359436, "num_tokens": 824186685.0, "step": 21600 }, { "epoch": 2.747869227833609, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.429490089416504, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8692487478256226, "num_tokens": 824223716.0, "step": 21601 }, { "epoch": 2.7479964381121995, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.48740005493164, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8733487725257874, "num_tokens": 824264627.0, "step": 21602 }, { "epoch": 2.74812364839079, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.394560813903809, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8746882677078247, "num_tokens": 824301901.0, "step": 21603 }, { "epoch": 2.7482508586693806, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.502058982849121, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8775984048843384, "num_tokens": 824344621.0, "step": 21604 }, { "epoch": 2.748378068947971, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.364455223083496, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8794633150100708, "num_tokens": 824385850.0, "step": 21605 }, { "epoch": 2.7485052792265616, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.417210578918457, "learning_rate": 1e-06, "loss": 0.313, "mean_token_accuracy": 0.9011757373809814, "num_tokens": 824425545.0, "step": 21606 }, { "epoch": 2.748632489505152, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.421950340270996, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8914991617202759, "num_tokens": 824468611.0, "step": 21607 }, { "epoch": 2.7487596997837427, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.449886322021484, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.87921142578125, "num_tokens": 824512161.0, "step": 21608 }, { "epoch": 2.7488869100623328, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.43310260772705, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8874030113220215, "num_tokens": 824555502.0, "step": 21609 }, { "epoch": 2.7490141203409237, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.471786499023438, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8681656122207642, "num_tokens": 824596277.0, "step": 21610 }, { "epoch": 2.749141330619514, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.464015007019043, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8634282946586609, "num_tokens": 824633704.0, "step": 21611 }, { "epoch": 2.749268540898105, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.42873477935791, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8791291117668152, "num_tokens": 824673584.0, "step": 21612 }, { "epoch": 2.749395751176695, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.516881942749023, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8691446781158447, "num_tokens": 824712508.0, "step": 21613 }, { "epoch": 2.749522961455286, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.444083213806152, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.885247528553009, "num_tokens": 824754353.0, "step": 21614 }, { "epoch": 2.749650171733876, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.513392448425293, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8558003902435303, "num_tokens": 824796333.0, "step": 21615 }, { "epoch": 2.7497773820124665, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.486394882202148, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8681414127349854, "num_tokens": 824832048.0, "step": 21616 }, { "epoch": 2.749904592291057, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.42215633392334, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8699920177459717, "num_tokens": 824871889.0, "step": 21617 }, { "epoch": 2.7500318025696475, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.50734806060791, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8849337100982666, "num_tokens": 824912547.0, "step": 21618 }, { "epoch": 2.750159012848238, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.46097469329834, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8552502393722534, "num_tokens": 824955626.0, "step": 21619 }, { "epoch": 2.7502862231268286, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.434481620788574, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8828218579292297, "num_tokens": 824991813.0, "step": 21620 }, { "epoch": 2.750413433405419, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.490813255310059, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8827256560325623, "num_tokens": 825024356.0, "step": 21621 }, { "epoch": 2.7505406436840096, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.449414253234863, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.872651994228363, "num_tokens": 825060316.0, "step": 21622 }, { "epoch": 2.7506678539626, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.541242599487305, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8672509789466858, "num_tokens": 825096044.0, "step": 21623 }, { "epoch": 2.7507950642411907, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.463823318481445, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8722728490829468, "num_tokens": 825123999.0, "step": 21624 }, { "epoch": 2.750922274519781, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.467236518859863, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8650405406951904, "num_tokens": 825170155.0, "step": 21625 }, { "epoch": 2.7510494847983717, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.493090629577637, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8741921782493591, "num_tokens": 825213388.0, "step": 21626 }, { "epoch": 2.7511766950769623, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.482880592346191, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.881935715675354, "num_tokens": 825247171.0, "step": 21627 }, { "epoch": 2.751303905355553, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.55115032196045, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8763633966445923, "num_tokens": 825283271.0, "step": 21628 }, { "epoch": 2.7514311156341433, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.419885635375977, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8738787174224854, "num_tokens": 825318976.0, "step": 21629 }, { "epoch": 2.751558325912734, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.5081148147583, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8742629289627075, "num_tokens": 825358019.0, "step": 21630 }, { "epoch": 2.7516855361913244, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.479194641113281, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8912991285324097, "num_tokens": 825394025.0, "step": 21631 }, { "epoch": 2.751812746469915, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.49269962310791, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8854103088378906, "num_tokens": 825433021.0, "step": 21632 }, { "epoch": 2.7519399567485054, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.488899230957031, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.875796914100647, "num_tokens": 825477026.0, "step": 21633 }, { "epoch": 2.7520671670270955, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.478700637817383, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8792657852172852, "num_tokens": 825511054.0, "step": 21634 }, { "epoch": 2.7521943773056865, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.495357513427734, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8607457280158997, "num_tokens": 825549654.0, "step": 21635 }, { "epoch": 2.7523215875842766, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.55759048461914, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8765461444854736, "num_tokens": 825587269.0, "step": 21636 }, { "epoch": 2.7524487978628676, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.385727882385254, "learning_rate": 1e-06, "loss": 0.3253, "mean_token_accuracy": 0.8977215886116028, "num_tokens": 825624864.0, "step": 21637 }, { "epoch": 2.7525760081414576, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.463072776794434, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8674364686012268, "num_tokens": 825667025.0, "step": 21638 }, { "epoch": 2.7527032184200486, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.397600173950195, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8600653409957886, "num_tokens": 825709505.0, "step": 21639 }, { "epoch": 2.7528304286986387, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.450177192687988, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8782860040664673, "num_tokens": 825747376.0, "step": 21640 }, { "epoch": 2.7529576389772292, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.578007698059082, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8811761736869812, "num_tokens": 825784817.0, "step": 21641 }, { "epoch": 2.7530848492558198, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.477627754211426, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.873142659664154, "num_tokens": 825815631.0, "step": 21642 }, { "epoch": 2.7532120595344103, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.421561241149902, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.874176025390625, "num_tokens": 825854202.0, "step": 21643 }, { "epoch": 2.753339269813001, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.489130020141602, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8835741281509399, "num_tokens": 825890636.0, "step": 21644 }, { "epoch": 2.7534664800915913, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.54672908782959, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8523479700088501, "num_tokens": 825931054.0, "step": 21645 }, { "epoch": 2.753593690370182, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.417814254760742, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8704633712768555, "num_tokens": 825972519.0, "step": 21646 }, { "epoch": 2.7537209006487724, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.498644828796387, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8763548731803894, "num_tokens": 826010010.0, "step": 21647 }, { "epoch": 2.753848110927363, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.43480110168457, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8745614886283875, "num_tokens": 826048348.0, "step": 21648 }, { "epoch": 2.7539753212059535, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.494140625, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8716083765029907, "num_tokens": 826087807.0, "step": 21649 }, { "epoch": 2.754102531484544, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.449080467224121, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8713415265083313, "num_tokens": 826124657.0, "step": 21650 }, { "epoch": 2.7542297417631345, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.516355514526367, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8773770928382874, "num_tokens": 826164190.0, "step": 21651 }, { "epoch": 2.754356952041725, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.507232666015625, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8758480548858643, "num_tokens": 826202876.0, "step": 21652 }, { "epoch": 2.7544841623203156, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.467257499694824, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8736773133277893, "num_tokens": 826242955.0, "step": 21653 }, { "epoch": 2.754611372598906, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.49052619934082, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8831809759140015, "num_tokens": 826279925.0, "step": 21654 }, { "epoch": 2.7547385828774966, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.50135612487793, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8853412866592407, "num_tokens": 826314076.0, "step": 21655 }, { "epoch": 2.754865793156087, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.484493255615234, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8714425563812256, "num_tokens": 826354138.0, "step": 21656 }, { "epoch": 2.7549930034346777, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.462469100952148, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8701847791671753, "num_tokens": 826395393.0, "step": 21657 }, { "epoch": 2.755120213713268, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.48397445678711, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8652905821800232, "num_tokens": 826437539.0, "step": 21658 }, { "epoch": 2.7552474239918583, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.483352661132812, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8685160279273987, "num_tokens": 826475857.0, "step": 21659 }, { "epoch": 2.7553746342704493, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.479899406433105, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8834654688835144, "num_tokens": 826513956.0, "step": 21660 }, { "epoch": 2.7555018445490393, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.454139709472656, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8731571435928345, "num_tokens": 826552747.0, "step": 21661 }, { "epoch": 2.7556290548276303, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.483113288879395, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8818728923797607, "num_tokens": 826590830.0, "step": 21662 }, { "epoch": 2.7557562651062204, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.497159004211426, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.877263069152832, "num_tokens": 826633414.0, "step": 21663 }, { "epoch": 2.755883475384811, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.481082916259766, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8840707540512085, "num_tokens": 826668344.0, "step": 21664 }, { "epoch": 2.7560106856634015, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.520169258117676, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8581815958023071, "num_tokens": 826710587.0, "step": 21665 }, { "epoch": 2.756137895941992, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.4008150100708, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8658528327941895, "num_tokens": 826748854.0, "step": 21666 }, { "epoch": 2.7562651062205825, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.521833419799805, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8616631031036377, "num_tokens": 826789591.0, "step": 21667 }, { "epoch": 2.756392316499173, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.509156227111816, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.871133029460907, "num_tokens": 826831890.0, "step": 21668 }, { "epoch": 2.7565195267777636, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.47543716430664, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8567663431167603, "num_tokens": 826870265.0, "step": 21669 }, { "epoch": 2.756646737056354, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.47896957397461, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8736286759376526, "num_tokens": 826910993.0, "step": 21670 }, { "epoch": 2.7567739473349446, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45426082611084, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8729591965675354, "num_tokens": 826947492.0, "step": 21671 }, { "epoch": 2.756901157613535, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.493692398071289, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8771539926528931, "num_tokens": 826987215.0, "step": 21672 }, { "epoch": 2.7570283678921257, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.5719575881958, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8716518878936768, "num_tokens": 827024092.0, "step": 21673 }, { "epoch": 2.757155578170716, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.438345909118652, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8574706315994263, "num_tokens": 827060541.0, "step": 21674 }, { "epoch": 2.7572827884493067, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.458951950073242, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8727895021438599, "num_tokens": 827101205.0, "step": 21675 }, { "epoch": 2.7574099987278973, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.421650886535645, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8679376840591431, "num_tokens": 827132441.0, "step": 21676 }, { "epoch": 2.757537209006488, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.561347007751465, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8750353455543518, "num_tokens": 827169530.0, "step": 21677 }, { "epoch": 2.7576644192850783, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.403589248657227, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8805383443832397, "num_tokens": 827207352.0, "step": 21678 }, { "epoch": 2.757791629563669, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.517454147338867, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8685775995254517, "num_tokens": 827242790.0, "step": 21679 }, { "epoch": 2.7579188398422594, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.490753173828125, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8758457899093628, "num_tokens": 827281676.0, "step": 21680 }, { "epoch": 2.75804605012085, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.445343971252441, "learning_rate": 1e-06, "loss": 0.3502, "mean_token_accuracy": 0.8880844116210938, "num_tokens": 827313676.0, "step": 21681 }, { "epoch": 2.7581732603994404, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.514589309692383, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8743347525596619, "num_tokens": 827355017.0, "step": 21682 }, { "epoch": 2.758300470678031, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.461604118347168, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8915737867355347, "num_tokens": 827391773.0, "step": 21683 }, { "epoch": 2.758427680956621, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.5333833694458, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8888769149780273, "num_tokens": 827433034.0, "step": 21684 }, { "epoch": 2.758554891235212, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.554357528686523, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8852605819702148, "num_tokens": 827475729.0, "step": 21685 }, { "epoch": 2.758682101513802, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.433877944946289, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8797922730445862, "num_tokens": 827514127.0, "step": 21686 }, { "epoch": 2.758809311792393, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.51307201385498, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8845375180244446, "num_tokens": 827546709.0, "step": 21687 }, { "epoch": 2.758936522070983, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.495504379272461, "learning_rate": 1e-06, "loss": 0.3335, "mean_token_accuracy": 0.8943167328834534, "num_tokens": 827581450.0, "step": 21688 }, { "epoch": 2.7590637323495737, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.44117259979248, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8709832429885864, "num_tokens": 827623161.0, "step": 21689 }, { "epoch": 2.7591909426281642, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.553211212158203, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8705066442489624, "num_tokens": 827659460.0, "step": 21690 }, { "epoch": 2.7593181529067548, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.426088333129883, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8694459199905396, "num_tokens": 827701110.0, "step": 21691 }, { "epoch": 2.7594453631853453, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45413589477539, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8715007305145264, "num_tokens": 827738992.0, "step": 21692 }, { "epoch": 2.759572573463936, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.603174209594727, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.859900176525116, "num_tokens": 827775853.0, "step": 21693 }, { "epoch": 2.7596997837425263, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.393671989440918, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8763014674186707, "num_tokens": 827819438.0, "step": 21694 }, { "epoch": 2.759826994021117, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.558929443359375, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8683848977088928, "num_tokens": 827858833.0, "step": 21695 }, { "epoch": 2.7599542042997074, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.501323699951172, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8682737946510315, "num_tokens": 827895069.0, "step": 21696 }, { "epoch": 2.760081414578298, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.554054260253906, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8878418803215027, "num_tokens": 827932970.0, "step": 21697 }, { "epoch": 2.7602086248568884, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.526775360107422, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.886518657207489, "num_tokens": 827981859.0, "step": 21698 }, { "epoch": 2.760335835135479, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.427510261535645, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8686625957489014, "num_tokens": 828022018.0, "step": 21699 }, { "epoch": 2.7604630454140695, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.513982772827148, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.87006676197052, "num_tokens": 828061546.0, "step": 21700 }, { "epoch": 2.76059025569266, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.497754096984863, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8608068227767944, "num_tokens": 828104464.0, "step": 21701 }, { "epoch": 2.7607174659712506, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.530766487121582, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8640714883804321, "num_tokens": 828142527.0, "step": 21702 }, { "epoch": 2.760844676249841, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.437808990478516, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8746912479400635, "num_tokens": 828189670.0, "step": 21703 }, { "epoch": 2.7609718865284316, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.464529037475586, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8813230991363525, "num_tokens": 828231511.0, "step": 21704 }, { "epoch": 2.761099096807022, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.468050003051758, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8738090395927429, "num_tokens": 828274387.0, "step": 21705 }, { "epoch": 2.7612263070856127, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.412665367126465, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8673326969146729, "num_tokens": 828312263.0, "step": 21706 }, { "epoch": 2.7613535173642028, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.523249626159668, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8753982782363892, "num_tokens": 828345550.0, "step": 21707 }, { "epoch": 2.7614807276427937, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.492016792297363, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8635355234146118, "num_tokens": 828378633.0, "step": 21708 }, { "epoch": 2.761607937921384, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.547465324401855, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.868198037147522, "num_tokens": 828414292.0, "step": 21709 }, { "epoch": 2.761735148199975, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.484030723571777, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8734864592552185, "num_tokens": 828452682.0, "step": 21710 }, { "epoch": 2.761862358478565, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.462389945983887, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8897727131843567, "num_tokens": 828494162.0, "step": 21711 }, { "epoch": 2.761989568757156, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.552952766418457, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8787193894386292, "num_tokens": 828525997.0, "step": 21712 }, { "epoch": 2.762116779035746, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.434918403625488, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8736364841461182, "num_tokens": 828562488.0, "step": 21713 }, { "epoch": 2.7622439893143365, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.524819374084473, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.8927376866340637, "num_tokens": 828591818.0, "step": 21714 }, { "epoch": 2.762371199592927, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.417929649353027, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8631879687309265, "num_tokens": 828628872.0, "step": 21715 }, { "epoch": 2.7624984098715175, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.499495506286621, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8833147883415222, "num_tokens": 828670939.0, "step": 21716 }, { "epoch": 2.762625620150108, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.585718154907227, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8572590351104736, "num_tokens": 828708446.0, "step": 21717 }, { "epoch": 2.7627528304286986, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.414801597595215, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8778348565101624, "num_tokens": 828749873.0, "step": 21718 }, { "epoch": 2.762880040707289, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.460368156433105, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8672075867652893, "num_tokens": 828790066.0, "step": 21719 }, { "epoch": 2.7630072509858796, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.546899795532227, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8714995384216309, "num_tokens": 828821911.0, "step": 21720 }, { "epoch": 2.76313446126447, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.487464904785156, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8693720698356628, "num_tokens": 828856572.0, "step": 21721 }, { "epoch": 2.7632616715430607, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.400873184204102, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8857678174972534, "num_tokens": 828898865.0, "step": 21722 }, { "epoch": 2.763388881821651, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.475850105285645, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8814820647239685, "num_tokens": 828933768.0, "step": 21723 }, { "epoch": 2.7635160921002417, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.499425888061523, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8896641135215759, "num_tokens": 828968869.0, "step": 21724 }, { "epoch": 2.7636433023788323, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.456807136535645, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8580604791641235, "num_tokens": 829005621.0, "step": 21725 }, { "epoch": 2.763770512657423, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.507489204406738, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8645855188369751, "num_tokens": 829045046.0, "step": 21726 }, { "epoch": 2.7638977229360133, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.491252899169922, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8870142698287964, "num_tokens": 829084588.0, "step": 21727 }, { "epoch": 2.764024933214604, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.580880165100098, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.85850989818573, "num_tokens": 829121010.0, "step": 21728 }, { "epoch": 2.7641521434931944, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.484424591064453, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8832859396934509, "num_tokens": 829159618.0, "step": 21729 }, { "epoch": 2.764279353771785, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.548713684082031, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8624314069747925, "num_tokens": 829204826.0, "step": 21730 }, { "epoch": 2.7644065640503754, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.550521850585938, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8852370977401733, "num_tokens": 829240741.0, "step": 21731 }, { "epoch": 2.7645337743289655, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.465808868408203, "learning_rate": 1e-06, "loss": 0.3463, "mean_token_accuracy": 0.8928022980690002, "num_tokens": 829278065.0, "step": 21732 }, { "epoch": 2.7646609846075565, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.533647537231445, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8707502484321594, "num_tokens": 829319438.0, "step": 21733 }, { "epoch": 2.7647881948861466, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.503388404846191, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8837465643882751, "num_tokens": 829357299.0, "step": 21734 }, { "epoch": 2.7649154051647375, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.512940406799316, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8756695985794067, "num_tokens": 829398600.0, "step": 21735 }, { "epoch": 2.7650426154433276, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.531375885009766, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8634307980537415, "num_tokens": 829439214.0, "step": 21736 }, { "epoch": 2.7651698257219186, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.499231338500977, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8639052510261536, "num_tokens": 829477634.0, "step": 21737 }, { "epoch": 2.7652970360005087, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.477566719055176, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8813031911849976, "num_tokens": 829514527.0, "step": 21738 }, { "epoch": 2.765424246279099, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.464407920837402, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8758927583694458, "num_tokens": 829558531.0, "step": 21739 }, { "epoch": 2.7655514565576897, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.517911911010742, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8775557279586792, "num_tokens": 829600769.0, "step": 21740 }, { "epoch": 2.7656786668362803, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.525980949401855, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8790081143379211, "num_tokens": 829635938.0, "step": 21741 }, { "epoch": 2.765805877114871, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.47148609161377, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8752203583717346, "num_tokens": 829672956.0, "step": 21742 }, { "epoch": 2.7659330873934613, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.56096076965332, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8741171360015869, "num_tokens": 829712658.0, "step": 21743 }, { "epoch": 2.766060297672052, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.482020378112793, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.877610981464386, "num_tokens": 829755429.0, "step": 21744 }, { "epoch": 2.7661875079506424, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.545981407165527, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8739397525787354, "num_tokens": 829796373.0, "step": 21745 }, { "epoch": 2.766314718229233, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.485320091247559, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8703013062477112, "num_tokens": 829834049.0, "step": 21746 }, { "epoch": 2.7664419285078234, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.48736572265625, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8744430541992188, "num_tokens": 829870853.0, "step": 21747 }, { "epoch": 2.766569138786414, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.482671737670898, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8680298328399658, "num_tokens": 829907519.0, "step": 21748 }, { "epoch": 2.7666963490650045, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.51695442199707, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8616690635681152, "num_tokens": 829947162.0, "step": 21749 }, { "epoch": 2.766823559343595, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.492100715637207, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8725329637527466, "num_tokens": 829991981.0, "step": 21750 }, { "epoch": 2.7669507696221856, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.447966575622559, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8804371356964111, "num_tokens": 830032409.0, "step": 21751 }, { "epoch": 2.767077979900776, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.482536315917969, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8772848844528198, "num_tokens": 830072738.0, "step": 21752 }, { "epoch": 2.7672051901793666, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.435269355773926, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8714078664779663, "num_tokens": 830110880.0, "step": 21753 }, { "epoch": 2.767332400457957, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.462346076965332, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8698693513870239, "num_tokens": 830150148.0, "step": 21754 }, { "epoch": 2.7674596107365477, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.439798355102539, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8628016710281372, "num_tokens": 830190717.0, "step": 21755 }, { "epoch": 2.767586821015138, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.393556594848633, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8756786584854126, "num_tokens": 830226972.0, "step": 21756 }, { "epoch": 2.7677140312937283, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.466672897338867, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.882400631904602, "num_tokens": 830262612.0, "step": 21757 }, { "epoch": 2.7678412415723193, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.428126335144043, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8805299401283264, "num_tokens": 830306856.0, "step": 21758 }, { "epoch": 2.7679684518509093, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.459173202514648, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8725959062576294, "num_tokens": 830340726.0, "step": 21759 }, { "epoch": 2.7680956621295003, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.472835540771484, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8742278814315796, "num_tokens": 830380697.0, "step": 21760 }, { "epoch": 2.7682228724080904, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.411293983459473, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8522012233734131, "num_tokens": 830414737.0, "step": 21761 }, { "epoch": 2.768350082686681, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.4296236038208, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8744124174118042, "num_tokens": 830456518.0, "step": 21762 }, { "epoch": 2.7684772929652715, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.488914489746094, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8707437515258789, "num_tokens": 830494192.0, "step": 21763 }, { "epoch": 2.768604503243862, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.446731567382812, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8677375316619873, "num_tokens": 830533659.0, "step": 21764 }, { "epoch": 2.7687317135224525, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.500223159790039, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8687608242034912, "num_tokens": 830569605.0, "step": 21765 }, { "epoch": 2.768858923801043, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.458290100097656, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8531529307365417, "num_tokens": 830610814.0, "step": 21766 }, { "epoch": 2.7689861340796336, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.444941520690918, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8734192848205566, "num_tokens": 830646811.0, "step": 21767 }, { "epoch": 2.769113344358224, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.485538482666016, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8774027824401855, "num_tokens": 830680984.0, "step": 21768 }, { "epoch": 2.7692405546368146, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.546826362609863, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8737623691558838, "num_tokens": 830713384.0, "step": 21769 }, { "epoch": 2.769367764915405, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.496475219726562, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8613600134849548, "num_tokens": 830746922.0, "step": 21770 }, { "epoch": 2.7694949751939957, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.43809986114502, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8797996640205383, "num_tokens": 830789459.0, "step": 21771 }, { "epoch": 2.769622185472586, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.500954627990723, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8711119294166565, "num_tokens": 830826736.0, "step": 21772 }, { "epoch": 2.7697493957511767, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.457029342651367, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.876708984375, "num_tokens": 830868118.0, "step": 21773 }, { "epoch": 2.7698766060297673, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.420634269714355, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8711156845092773, "num_tokens": 830911658.0, "step": 21774 }, { "epoch": 2.770003816308358, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.487265586853027, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.875441312789917, "num_tokens": 830948456.0, "step": 21775 }, { "epoch": 2.7701310265869483, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.400762557983398, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8829733729362488, "num_tokens": 830988790.0, "step": 21776 }, { "epoch": 2.770258236865539, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.52906608581543, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8671329021453857, "num_tokens": 831024551.0, "step": 21777 }, { "epoch": 2.7703854471441294, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.453367233276367, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8709882497787476, "num_tokens": 831060082.0, "step": 21778 }, { "epoch": 2.77051265742272, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.465860366821289, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8752076625823975, "num_tokens": 831100067.0, "step": 21779 }, { "epoch": 2.77063986770131, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.467568397521973, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8847924470901489, "num_tokens": 831131923.0, "step": 21780 }, { "epoch": 2.770767077979901, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.467826843261719, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8595527410507202, "num_tokens": 831167985.0, "step": 21781 }, { "epoch": 2.770894288258491, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.449010848999023, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8688353300094604, "num_tokens": 831211737.0, "step": 21782 }, { "epoch": 2.771021498537082, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.509286880493164, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8634279370307922, "num_tokens": 831248764.0, "step": 21783 }, { "epoch": 2.771148708815672, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45418643951416, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8797270059585571, "num_tokens": 831284140.0, "step": 21784 }, { "epoch": 2.771275919094263, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.533517837524414, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8770322203636169, "num_tokens": 831318567.0, "step": 21785 }, { "epoch": 2.771403129372853, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.43593978881836, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8672767877578735, "num_tokens": 831357128.0, "step": 21786 }, { "epoch": 2.7715303396514437, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.521167755126953, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8536508083343506, "num_tokens": 831391551.0, "step": 21787 }, { "epoch": 2.771657549930034, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.510689735412598, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8694683313369751, "num_tokens": 831430332.0, "step": 21788 }, { "epoch": 2.7717847602086247, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.417906761169434, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8790404796600342, "num_tokens": 831471639.0, "step": 21789 }, { "epoch": 2.7719119704872153, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.578691482543945, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8788475394248962, "num_tokens": 831506470.0, "step": 21790 }, { "epoch": 2.772039180765806, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.427719116210938, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8675446510314941, "num_tokens": 831547212.0, "step": 21791 }, { "epoch": 2.7721663910443963, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.462885856628418, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8586186170578003, "num_tokens": 831593181.0, "step": 21792 }, { "epoch": 2.772293601322987, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.534466743469238, "learning_rate": 1e-06, "loss": 0.3412, "mean_token_accuracy": 0.8892923593521118, "num_tokens": 831624889.0, "step": 21793 }, { "epoch": 2.7724208116015774, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.486038208007812, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8801230788230896, "num_tokens": 831662389.0, "step": 21794 }, { "epoch": 2.772548021880168, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.423720359802246, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.875933051109314, "num_tokens": 831705282.0, "step": 21795 }, { "epoch": 2.7726752321587584, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.458207130432129, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8807525038719177, "num_tokens": 831741604.0, "step": 21796 }, { "epoch": 2.772802442437349, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.430930137634277, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8652315735816956, "num_tokens": 831784165.0, "step": 21797 }, { "epoch": 2.7729296527159395, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.56922435760498, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8789747953414917, "num_tokens": 831821101.0, "step": 21798 }, { "epoch": 2.77305686299453, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.451034545898438, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8758108615875244, "num_tokens": 831855766.0, "step": 21799 }, { "epoch": 2.7731840732731206, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.502756118774414, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8792887926101685, "num_tokens": 831895692.0, "step": 21800 }, { "epoch": 2.773311283551711, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.508014678955078, "learning_rate": 1e-06, "loss": 0.3331, "mean_token_accuracy": 0.892775297164917, "num_tokens": 831931079.0, "step": 21801 }, { "epoch": 2.7734384938303016, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.500816345214844, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.8490140438079834, "num_tokens": 831969742.0, "step": 21802 }, { "epoch": 2.773565704108892, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.521449089050293, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8636279702186584, "num_tokens": 832011059.0, "step": 21803 }, { "epoch": 2.7736929143874827, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.48864459991455, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.883111834526062, "num_tokens": 832042768.0, "step": 21804 }, { "epoch": 2.7738201246660728, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.486406326293945, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8820095658302307, "num_tokens": 832077699.0, "step": 21805 }, { "epoch": 2.7739473349446637, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.380654335021973, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8822058439254761, "num_tokens": 832113528.0, "step": 21806 }, { "epoch": 2.774074545223254, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.474163055419922, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8782389163970947, "num_tokens": 832149108.0, "step": 21807 }, { "epoch": 2.774201755501845, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.465880393981934, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.891289234161377, "num_tokens": 832189279.0, "step": 21808 }, { "epoch": 2.774328965780435, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45897388458252, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8756141662597656, "num_tokens": 832224423.0, "step": 21809 }, { "epoch": 2.774456176059026, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55140495300293, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8679615259170532, "num_tokens": 832260389.0, "step": 21810 }, { "epoch": 2.774583386337616, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.53748893737793, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8735306262969971, "num_tokens": 832297591.0, "step": 21811 }, { "epoch": 2.7747105966162064, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.495264053344727, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8792667984962463, "num_tokens": 832330195.0, "step": 21812 }, { "epoch": 2.774837806894797, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.444616317749023, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8816989660263062, "num_tokens": 832361613.0, "step": 21813 }, { "epoch": 2.7749650171733875, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.436568260192871, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8802774548530579, "num_tokens": 832402800.0, "step": 21814 }, { "epoch": 2.775092227451978, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.443754196166992, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8764104843139648, "num_tokens": 832445161.0, "step": 21815 }, { "epoch": 2.7752194377305686, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.528903007507324, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8592365384101868, "num_tokens": 832487905.0, "step": 21816 }, { "epoch": 2.775346648009159, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.493789672851562, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8760522603988647, "num_tokens": 832521107.0, "step": 21817 }, { "epoch": 2.7754738582877496, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.495254516601562, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8702086210250854, "num_tokens": 832557213.0, "step": 21818 }, { "epoch": 2.77560106856634, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.485515594482422, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8619827032089233, "num_tokens": 832600445.0, "step": 21819 }, { "epoch": 2.7757282788449307, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.463799476623535, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.873530924320221, "num_tokens": 832635994.0, "step": 21820 }, { "epoch": 2.775855489123521, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.472349166870117, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8853250741958618, "num_tokens": 832677151.0, "step": 21821 }, { "epoch": 2.7759826994021117, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.432024955749512, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8734071850776672, "num_tokens": 832713995.0, "step": 21822 }, { "epoch": 2.7761099096807023, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.529341697692871, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8650900721549988, "num_tokens": 832754201.0, "step": 21823 }, { "epoch": 2.776237119959293, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.443947792053223, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8711745738983154, "num_tokens": 832796920.0, "step": 21824 }, { "epoch": 2.7763643302378833, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.432968139648438, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8710019588470459, "num_tokens": 832835952.0, "step": 21825 }, { "epoch": 2.776491540516474, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.480889320373535, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8766208291053772, "num_tokens": 832878306.0, "step": 21826 }, { "epoch": 2.7766187507950644, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.53948974609375, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8668196201324463, "num_tokens": 832914315.0, "step": 21827 }, { "epoch": 2.776745961073655, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.487056732177734, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8701735734939575, "num_tokens": 832956661.0, "step": 21828 }, { "epoch": 2.7768731713522454, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.417765617370605, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8800868988037109, "num_tokens": 832999446.0, "step": 21829 }, { "epoch": 2.7770003816308355, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.531752586364746, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8476094603538513, "num_tokens": 833039001.0, "step": 21830 }, { "epoch": 2.7771275919094265, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.418292999267578, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8727000951766968, "num_tokens": 833074334.0, "step": 21831 }, { "epoch": 2.7772548021880166, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.455737113952637, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8621654510498047, "num_tokens": 833113089.0, "step": 21832 }, { "epoch": 2.7773820124666075, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.499752044677734, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8710800409317017, "num_tokens": 833149755.0, "step": 21833 }, { "epoch": 2.7775092227451976, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.441983222961426, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8649507761001587, "num_tokens": 833187718.0, "step": 21834 }, { "epoch": 2.7776364330237886, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.430006980895996, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8811624646186829, "num_tokens": 833221416.0, "step": 21835 }, { "epoch": 2.7777636433023787, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.47260856628418, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8875463008880615, "num_tokens": 833261101.0, "step": 21836 }, { "epoch": 2.777890853580969, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.466144561767578, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8822949528694153, "num_tokens": 833296804.0, "step": 21837 }, { "epoch": 2.7780180638595597, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.476614952087402, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8748661875724792, "num_tokens": 833332005.0, "step": 21838 }, { "epoch": 2.7781452741381503, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.438800811767578, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8701545000076294, "num_tokens": 833371991.0, "step": 21839 }, { "epoch": 2.778272484416741, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.444744110107422, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8856748342514038, "num_tokens": 833407514.0, "step": 21840 }, { "epoch": 2.7783996946953313, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.423656463623047, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8719688057899475, "num_tokens": 833444519.0, "step": 21841 }, { "epoch": 2.778526904973922, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.507938385009766, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8693943023681641, "num_tokens": 833482252.0, "step": 21842 }, { "epoch": 2.7786541152525124, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45625114440918, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8719493746757507, "num_tokens": 833515513.0, "step": 21843 }, { "epoch": 2.778781325531103, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.43115234375, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8812159299850464, "num_tokens": 833551188.0, "step": 21844 }, { "epoch": 2.7789085358096934, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.476222038269043, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8831015825271606, "num_tokens": 833589159.0, "step": 21845 }, { "epoch": 2.779035746088284, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.4774808883667, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8646713495254517, "num_tokens": 833623581.0, "step": 21846 }, { "epoch": 2.7791629563668745, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.45903491973877, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.862765908241272, "num_tokens": 833660466.0, "step": 21847 }, { "epoch": 2.779290166645465, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.460432052612305, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8636759519577026, "num_tokens": 833697465.0, "step": 21848 }, { "epoch": 2.7794173769240555, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.447122573852539, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8716115951538086, "num_tokens": 833735178.0, "step": 21849 }, { "epoch": 2.779544587202646, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.504020690917969, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.8913184404373169, "num_tokens": 833770193.0, "step": 21850 }, { "epoch": 2.7796717974812366, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.404230117797852, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8740990161895752, "num_tokens": 833807836.0, "step": 21851 }, { "epoch": 2.779799007759827, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.481179237365723, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.870826005935669, "num_tokens": 833847605.0, "step": 21852 }, { "epoch": 2.7799262180384177, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.458000183105469, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8771804571151733, "num_tokens": 833884239.0, "step": 21853 }, { "epoch": 2.780053428317008, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.553919792175293, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8743777275085449, "num_tokens": 833921305.0, "step": 21854 }, { "epoch": 2.7801806385955983, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.49824047088623, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8648971319198608, "num_tokens": 833959297.0, "step": 21855 }, { "epoch": 2.7803078488741892, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.513901710510254, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8517698049545288, "num_tokens": 833992135.0, "step": 21856 }, { "epoch": 2.7804350591527793, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.502999305725098, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8645209074020386, "num_tokens": 834035933.0, "step": 21857 }, { "epoch": 2.7805622694313703, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.455303192138672, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8735926747322083, "num_tokens": 834070187.0, "step": 21858 }, { "epoch": 2.7806894797099604, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.527729988098145, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8750531673431396, "num_tokens": 834107370.0, "step": 21859 }, { "epoch": 2.780816689988551, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.511448860168457, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.865933358669281, "num_tokens": 834145357.0, "step": 21860 }, { "epoch": 2.7809439002671414, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.445162773132324, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8715976476669312, "num_tokens": 834187250.0, "step": 21861 }, { "epoch": 2.781071110545732, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.523551940917969, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.883929431438446, "num_tokens": 834227598.0, "step": 21862 }, { "epoch": 2.7811983208243225, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.45537281036377, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8886474370956421, "num_tokens": 834267032.0, "step": 21863 }, { "epoch": 2.781325531102913, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.51034927368164, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8630595207214355, "num_tokens": 834308479.0, "step": 21864 }, { "epoch": 2.7814527413815036, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.518320083618164, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8818647861480713, "num_tokens": 834346185.0, "step": 21865 }, { "epoch": 2.781579951660094, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.504044532775879, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8817811012268066, "num_tokens": 834384404.0, "step": 21866 }, { "epoch": 2.7817071619386846, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.4681396484375, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8718598484992981, "num_tokens": 834424833.0, "step": 21867 }, { "epoch": 2.781834372217275, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.47146224975586, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.876160204410553, "num_tokens": 834464215.0, "step": 21868 }, { "epoch": 2.7819615824958657, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.47720718383789, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8511124849319458, "num_tokens": 834502452.0, "step": 21869 }, { "epoch": 2.782088792774456, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.508020401000977, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8739932775497437, "num_tokens": 834539690.0, "step": 21870 }, { "epoch": 2.7822160030530467, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.41681957244873, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8638131618499756, "num_tokens": 834583193.0, "step": 21871 }, { "epoch": 2.7823432133316373, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.405656814575195, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8683990240097046, "num_tokens": 834620028.0, "step": 21872 }, { "epoch": 2.782470423610228, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.550832748413086, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.863766074180603, "num_tokens": 834657351.0, "step": 21873 }, { "epoch": 2.7825976338888183, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.486502647399902, "learning_rate": 1e-06, "loss": 0.3396, "mean_token_accuracy": 0.8876256346702576, "num_tokens": 834693687.0, "step": 21874 }, { "epoch": 2.782724844167409, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.474590301513672, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8727023601531982, "num_tokens": 834736171.0, "step": 21875 }, { "epoch": 2.7828520544459994, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.420500755310059, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8875815868377686, "num_tokens": 834775699.0, "step": 21876 }, { "epoch": 2.78297926472459, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.51755142211914, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8678577542304993, "num_tokens": 834816053.0, "step": 21877 }, { "epoch": 2.78310647500318, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.562687873840332, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8701289892196655, "num_tokens": 834857125.0, "step": 21878 }, { "epoch": 2.783233685281771, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.512447357177734, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8859837651252747, "num_tokens": 834895676.0, "step": 21879 }, { "epoch": 2.783360895560361, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.586732864379883, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8771798610687256, "num_tokens": 834927748.0, "step": 21880 }, { "epoch": 2.783488105838952, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.480669021606445, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.869973361492157, "num_tokens": 834966722.0, "step": 21881 }, { "epoch": 2.783615316117542, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.493792533874512, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8751935958862305, "num_tokens": 835003770.0, "step": 21882 }, { "epoch": 2.783742526396133, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.49783992767334, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8821335434913635, "num_tokens": 835041504.0, "step": 21883 }, { "epoch": 2.783869736674723, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.540014266967773, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8751387596130371, "num_tokens": 835075179.0, "step": 21884 }, { "epoch": 2.7839969469533137, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.559126853942871, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8631336688995361, "num_tokens": 835113080.0, "step": 21885 }, { "epoch": 2.784124157231904, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.558196067810059, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8721928596496582, "num_tokens": 835150586.0, "step": 21886 }, { "epoch": 2.7842513675104947, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.484983444213867, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.871541440486908, "num_tokens": 835184049.0, "step": 21887 }, { "epoch": 2.7843785777890853, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.581076622009277, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8683833479881287, "num_tokens": 835222745.0, "step": 21888 }, { "epoch": 2.784505788067676, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.544928550720215, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.862017035484314, "num_tokens": 835263017.0, "step": 21889 }, { "epoch": 2.7846329983462663, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.43414306640625, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8872759938240051, "num_tokens": 835303916.0, "step": 21890 }, { "epoch": 2.784760208624857, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.545700073242188, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8883834481239319, "num_tokens": 835343148.0, "step": 21891 }, { "epoch": 2.7848874189034474, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.599416732788086, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8711206912994385, "num_tokens": 835382337.0, "step": 21892 }, { "epoch": 2.785014629182038, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.437146186828613, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8673578500747681, "num_tokens": 835423645.0, "step": 21893 }, { "epoch": 2.7851418394606284, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.5928316116333, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8466776609420776, "num_tokens": 835464590.0, "step": 21894 }, { "epoch": 2.785269049739219, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.454673767089844, "learning_rate": 1e-06, "loss": 0.3333, "mean_token_accuracy": 0.8964480757713318, "num_tokens": 835496376.0, "step": 21895 }, { "epoch": 2.7853962600178095, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.623723983764648, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8583997488021851, "num_tokens": 835541231.0, "step": 21896 }, { "epoch": 2.7855234702964, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.507118225097656, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8779474496841431, "num_tokens": 835584038.0, "step": 21897 }, { "epoch": 2.7856506805749905, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.434505462646484, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8671674132347107, "num_tokens": 835627189.0, "step": 21898 }, { "epoch": 2.785777890853581, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.580059051513672, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8743237257003784, "num_tokens": 835659701.0, "step": 21899 }, { "epoch": 2.7859051011321716, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.44882583618164, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8797687888145447, "num_tokens": 835694045.0, "step": 21900 }, { "epoch": 2.786032311410762, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.526334762573242, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8709871172904968, "num_tokens": 835732118.0, "step": 21901 }, { "epoch": 2.7861595216893527, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.443581581115723, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8827065229415894, "num_tokens": 835772700.0, "step": 21902 }, { "epoch": 2.7862867319679427, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.569655418395996, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.857867419719696, "num_tokens": 835811206.0, "step": 21903 }, { "epoch": 2.7864139422465337, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.43945598602295, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8795361518859863, "num_tokens": 835844381.0, "step": 21904 }, { "epoch": 2.786541152525124, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.561448097229004, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8665419816970825, "num_tokens": 835879926.0, "step": 21905 }, { "epoch": 2.7866683628037148, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.455424308776855, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8632715344429016, "num_tokens": 835924014.0, "step": 21906 }, { "epoch": 2.786795573082305, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.44851303100586, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8665615320205688, "num_tokens": 835956054.0, "step": 21907 }, { "epoch": 2.786922783360896, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.47167682647705, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8801844120025635, "num_tokens": 835988885.0, "step": 21908 }, { "epoch": 2.787049993639486, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.472707748413086, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8718658685684204, "num_tokens": 836028885.0, "step": 21909 }, { "epoch": 2.7871772039180764, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.474176406860352, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8863313794136047, "num_tokens": 836067340.0, "step": 21910 }, { "epoch": 2.787304414196667, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.540319442749023, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8739296197891235, "num_tokens": 836104432.0, "step": 21911 }, { "epoch": 2.7874316244752575, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.482072830200195, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8689558506011963, "num_tokens": 836141197.0, "step": 21912 }, { "epoch": 2.787558834753848, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.493358612060547, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8851927518844604, "num_tokens": 836176949.0, "step": 21913 }, { "epoch": 2.7876860450324386, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.666431427001953, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8708723783493042, "num_tokens": 836206484.0, "step": 21914 }, { "epoch": 2.787813255311029, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.449427604675293, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8789242506027222, "num_tokens": 836243056.0, "step": 21915 }, { "epoch": 2.7879404655896196, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.42902946472168, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8687126636505127, "num_tokens": 836281154.0, "step": 21916 }, { "epoch": 2.78806767586821, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.569999694824219, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.87251216173172, "num_tokens": 836319895.0, "step": 21917 }, { "epoch": 2.7881948861468007, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.465058326721191, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.878955602645874, "num_tokens": 836357227.0, "step": 21918 }, { "epoch": 2.788322096425391, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.57064437866211, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8661306500434875, "num_tokens": 836398442.0, "step": 21919 }, { "epoch": 2.7884493067039817, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.558897018432617, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8818987607955933, "num_tokens": 836436660.0, "step": 21920 }, { "epoch": 2.7885765169825723, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.526656150817871, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8678030371665955, "num_tokens": 836479180.0, "step": 21921 }, { "epoch": 2.788703727261163, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.545892715454102, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8797034025192261, "num_tokens": 836518874.0, "step": 21922 }, { "epoch": 2.7888309375397533, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.512192726135254, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8702955842018127, "num_tokens": 836556910.0, "step": 21923 }, { "epoch": 2.788958147818344, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.579504013061523, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8855383396148682, "num_tokens": 836598151.0, "step": 21924 }, { "epoch": 2.7890853580969344, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.458098411560059, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8660036325454712, "num_tokens": 836637952.0, "step": 21925 }, { "epoch": 2.789212568375525, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.569906234741211, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8860098123550415, "num_tokens": 836677138.0, "step": 21926 }, { "epoch": 2.7893397786541154, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.451783180236816, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8785098791122437, "num_tokens": 836716880.0, "step": 21927 }, { "epoch": 2.7894669889327055, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.520648002624512, "learning_rate": 1e-06, "loss": 0.3582, "mean_token_accuracy": 0.8860491514205933, "num_tokens": 836752572.0, "step": 21928 }, { "epoch": 2.7895941992112965, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.472500801086426, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.869854211807251, "num_tokens": 836794321.0, "step": 21929 }, { "epoch": 2.7897214094898866, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.504633903503418, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8876941204071045, "num_tokens": 836829183.0, "step": 21930 }, { "epoch": 2.7898486197684775, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.53585433959961, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8623600602149963, "num_tokens": 836864644.0, "step": 21931 }, { "epoch": 2.7899758300470676, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.44766902923584, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8855016827583313, "num_tokens": 836895384.0, "step": 21932 }, { "epoch": 2.7901030403256586, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.522245407104492, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8760576248168945, "num_tokens": 836943336.0, "step": 21933 }, { "epoch": 2.7902302506042487, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.55339527130127, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8876984119415283, "num_tokens": 836980708.0, "step": 21934 }, { "epoch": 2.790357460882839, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.54561996459961, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8772484064102173, "num_tokens": 837015738.0, "step": 21935 }, { "epoch": 2.7904846711614297, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.544443130493164, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8796696662902832, "num_tokens": 837053014.0, "step": 21936 }, { "epoch": 2.7906118814400203, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.557671546936035, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8680678606033325, "num_tokens": 837083579.0, "step": 21937 }, { "epoch": 2.790739091718611, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.512550354003906, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8825554251670837, "num_tokens": 837126201.0, "step": 21938 }, { "epoch": 2.7908663019972013, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.576549530029297, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8833443522453308, "num_tokens": 837162400.0, "step": 21939 }, { "epoch": 2.790993512275792, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.480351448059082, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8782968521118164, "num_tokens": 837205733.0, "step": 21940 }, { "epoch": 2.7911207225543824, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.451261520385742, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8633496761322021, "num_tokens": 837243838.0, "step": 21941 }, { "epoch": 2.791247932832973, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.488504409790039, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8695394992828369, "num_tokens": 837280936.0, "step": 21942 }, { "epoch": 2.7913751431115634, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.516258239746094, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8737912178039551, "num_tokens": 837321790.0, "step": 21943 }, { "epoch": 2.791502353390154, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.496984481811523, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8768239617347717, "num_tokens": 837362023.0, "step": 21944 }, { "epoch": 2.7916295636687445, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.49083423614502, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8810491561889648, "num_tokens": 837404876.0, "step": 21945 }, { "epoch": 2.791756773947335, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.453933715820312, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.867190957069397, "num_tokens": 837442899.0, "step": 21946 }, { "epoch": 2.7918839842259255, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.522951126098633, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8833795785903931, "num_tokens": 837477611.0, "step": 21947 }, { "epoch": 2.792011194504516, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.497517585754395, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8615008592605591, "num_tokens": 837515928.0, "step": 21948 }, { "epoch": 2.7921384047831066, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.49355697631836, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8749962449073792, "num_tokens": 837553042.0, "step": 21949 }, { "epoch": 2.792265615061697, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.498968124389648, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8802520036697388, "num_tokens": 837589749.0, "step": 21950 }, { "epoch": 2.7923928253402877, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.411492347717285, "learning_rate": 1e-06, "loss": 0.3492, "mean_token_accuracy": 0.8920539617538452, "num_tokens": 837631104.0, "step": 21951 }, { "epoch": 2.792520035618878, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.546493530273438, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8830941915512085, "num_tokens": 837668173.0, "step": 21952 }, { "epoch": 2.7926472458974683, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.559812545776367, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8805217742919922, "num_tokens": 837704282.0, "step": 21953 }, { "epoch": 2.7927744561760592, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.482455253601074, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.864675760269165, "num_tokens": 837750790.0, "step": 21954 }, { "epoch": 2.7929016664546493, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.591043472290039, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8719354867935181, "num_tokens": 837792041.0, "step": 21955 }, { "epoch": 2.7930288767332403, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.569550514221191, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8798274993896484, "num_tokens": 837818645.0, "step": 21956 }, { "epoch": 2.7931560870118304, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.484769821166992, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8836176991462708, "num_tokens": 837859816.0, "step": 21957 }, { "epoch": 2.793283297290421, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.541176795959473, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8693451285362244, "num_tokens": 837903970.0, "step": 21958 }, { "epoch": 2.7934105075690114, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.508736610412598, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8739541172981262, "num_tokens": 837941158.0, "step": 21959 }, { "epoch": 2.793537717847602, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.56175422668457, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8798960447311401, "num_tokens": 837978157.0, "step": 21960 }, { "epoch": 2.7936649281261925, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.50916576385498, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8856160640716553, "num_tokens": 838017871.0, "step": 21961 }, { "epoch": 2.793792138404783, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.590514183044434, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8863890171051025, "num_tokens": 838050521.0, "step": 21962 }, { "epoch": 2.7939193486833735, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.576494216918945, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8814132213592529, "num_tokens": 838084450.0, "step": 21963 }, { "epoch": 2.794046558961964, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.546119689941406, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8691836595535278, "num_tokens": 838121505.0, "step": 21964 }, { "epoch": 2.7941737692405546, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.504611015319824, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8729211091995239, "num_tokens": 838166148.0, "step": 21965 }, { "epoch": 2.794300979519145, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.549849510192871, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8807557821273804, "num_tokens": 838203578.0, "step": 21966 }, { "epoch": 2.7944281897977357, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.513199806213379, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8800092935562134, "num_tokens": 838240702.0, "step": 21967 }, { "epoch": 2.794555400076326, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.553995132446289, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8816356658935547, "num_tokens": 838275639.0, "step": 21968 }, { "epoch": 2.7946826103549167, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.581302642822266, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8755447864532471, "num_tokens": 838310045.0, "step": 21969 }, { "epoch": 2.7948098206335072, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.481748580932617, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8753906488418579, "num_tokens": 838351571.0, "step": 21970 }, { "epoch": 2.7949370309120978, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.556317329406738, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8868058919906616, "num_tokens": 838393804.0, "step": 21971 }, { "epoch": 2.7950642411906883, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.439705848693848, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8744430541992188, "num_tokens": 838435516.0, "step": 21972 }, { "epoch": 2.795191451469279, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.525111198425293, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8819788694381714, "num_tokens": 838465425.0, "step": 21973 }, { "epoch": 2.7953186617478694, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.51201343536377, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8723650574684143, "num_tokens": 838512832.0, "step": 21974 }, { "epoch": 2.79544587202646, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.511496543884277, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8815087080001831, "num_tokens": 838552806.0, "step": 21975 }, { "epoch": 2.79557308230505, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.509660720825195, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8730117082595825, "num_tokens": 838588592.0, "step": 21976 }, { "epoch": 2.795700292583641, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.490530967712402, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8723539113998413, "num_tokens": 838629402.0, "step": 21977 }, { "epoch": 2.795827502862231, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.524770736694336, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8666481971740723, "num_tokens": 838676580.0, "step": 21978 }, { "epoch": 2.795954713140822, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.534147262573242, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8692836761474609, "num_tokens": 838715773.0, "step": 21979 }, { "epoch": 2.796081923419412, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.601255416870117, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8651283383369446, "num_tokens": 838750270.0, "step": 21980 }, { "epoch": 2.796209133698003, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.437609672546387, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8540648221969604, "num_tokens": 838794823.0, "step": 21981 }, { "epoch": 2.796336343976593, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.500472068786621, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8664738535881042, "num_tokens": 838828901.0, "step": 21982 }, { "epoch": 2.7964635542551837, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.509561538696289, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8663598299026489, "num_tokens": 838872648.0, "step": 21983 }, { "epoch": 2.796590764533774, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.501741409301758, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8872622847557068, "num_tokens": 838905150.0, "step": 21984 }, { "epoch": 2.7967179748123647, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.457005500793457, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8745548725128174, "num_tokens": 838949027.0, "step": 21985 }, { "epoch": 2.7968451850909553, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.488252639770508, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8759407997131348, "num_tokens": 838987320.0, "step": 21986 }, { "epoch": 2.796972395369546, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.43946647644043, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8701471090316772, "num_tokens": 839027396.0, "step": 21987 }, { "epoch": 2.7970996056481363, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.425312995910645, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.880477786064148, "num_tokens": 839069779.0, "step": 21988 }, { "epoch": 2.797226815926727, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.48436164855957, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8807815909385681, "num_tokens": 839110919.0, "step": 21989 }, { "epoch": 2.7973540262053174, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.550334930419922, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8885176181793213, "num_tokens": 839143689.0, "step": 21990 }, { "epoch": 2.797481236483908, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.413472175598145, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8563539981842041, "num_tokens": 839181274.0, "step": 21991 }, { "epoch": 2.7976084467624984, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.564592361450195, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8587620258331299, "num_tokens": 839217038.0, "step": 21992 }, { "epoch": 2.797735657041089, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.481337547302246, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.873808741569519, "num_tokens": 839260104.0, "step": 21993 }, { "epoch": 2.7978628673196795, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.491623878479004, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8775845170021057, "num_tokens": 839294328.0, "step": 21994 }, { "epoch": 2.79799007759827, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.468498229980469, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8736742734909058, "num_tokens": 839326662.0, "step": 21995 }, { "epoch": 2.7981172878768605, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.52875804901123, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8757824301719666, "num_tokens": 839355997.0, "step": 21996 }, { "epoch": 2.798244498155451, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.470327377319336, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8804829120635986, "num_tokens": 839396596.0, "step": 21997 }, { "epoch": 2.7983717084340416, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.443645477294922, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.875298023223877, "num_tokens": 839432251.0, "step": 21998 }, { "epoch": 2.798498918712632, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.455440521240234, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8832651376724243, "num_tokens": 839465987.0, "step": 21999 }, { "epoch": 2.7986261289912227, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.488746643066406, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8773208856582642, "num_tokens": 839506147.0, "step": 22000 }, { "epoch": 2.7987533392698127, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.516934394836426, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8661081194877625, "num_tokens": 839544620.0, "step": 22001 }, { "epoch": 2.7988805495484037, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.516486167907715, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8667823076248169, "num_tokens": 839583217.0, "step": 22002 }, { "epoch": 2.799007759826994, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.537054061889648, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8924956321716309, "num_tokens": 839621283.0, "step": 22003 }, { "epoch": 2.7991349701055848, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.529533386230469, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.870003342628479, "num_tokens": 839658072.0, "step": 22004 }, { "epoch": 2.799262180384175, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.457099914550781, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8577229380607605, "num_tokens": 839700926.0, "step": 22005 }, { "epoch": 2.799389390662766, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.527493476867676, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8835896849632263, "num_tokens": 839736160.0, "step": 22006 }, { "epoch": 2.799516600941356, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.495009422302246, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8892627358436584, "num_tokens": 839776834.0, "step": 22007 }, { "epoch": 2.7996438112199464, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.519307136535645, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8678755760192871, "num_tokens": 839818156.0, "step": 22008 }, { "epoch": 2.799771021498537, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.490910530090332, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8678950071334839, "num_tokens": 839857133.0, "step": 22009 }, { "epoch": 2.7998982317771275, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.551353454589844, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8592259287834167, "num_tokens": 839893129.0, "step": 22010 }, { "epoch": 2.800025442055718, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.497879028320312, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8777689933776855, "num_tokens": 839924622.0, "step": 22011 }, { "epoch": 2.8001526523343085, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.498930931091309, "learning_rate": 1e-06, "loss": 0.3418, "mean_token_accuracy": 0.8943286538124084, "num_tokens": 839961276.0, "step": 22012 }, { "epoch": 2.800279862612899, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.54367733001709, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8647683262825012, "num_tokens": 840002554.0, "step": 22013 }, { "epoch": 2.8004070728914896, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.482704162597656, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8723278045654297, "num_tokens": 840041642.0, "step": 22014 }, { "epoch": 2.80053428317008, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.466476440429688, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8825160264968872, "num_tokens": 840083532.0, "step": 22015 }, { "epoch": 2.8006614934486707, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.487712860107422, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8702419996261597, "num_tokens": 840124221.0, "step": 22016 }, { "epoch": 2.800788703727261, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.547102928161621, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8786460757255554, "num_tokens": 840162309.0, "step": 22017 }, { "epoch": 2.8009159140058517, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.484619140625, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8761513233184814, "num_tokens": 840194778.0, "step": 22018 }, { "epoch": 2.8010431242844422, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.537010192871094, "learning_rate": 1e-06, "loss": 0.342, "mean_token_accuracy": 0.8915747404098511, "num_tokens": 840225081.0, "step": 22019 }, { "epoch": 2.8011703345630328, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.457540512084961, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8736119866371155, "num_tokens": 840267698.0, "step": 22020 }, { "epoch": 2.8012975448416233, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.522710800170898, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8667641282081604, "num_tokens": 840303255.0, "step": 22021 }, { "epoch": 2.801424755120214, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.544079780578613, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8729287385940552, "num_tokens": 840344616.0, "step": 22022 }, { "epoch": 2.8015519653988044, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.452017784118652, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8756387233734131, "num_tokens": 840383293.0, "step": 22023 }, { "epoch": 2.801679175677395, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.539900779724121, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.864755392074585, "num_tokens": 840424583.0, "step": 22024 }, { "epoch": 2.8018063859559854, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.5241117477417, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.869074821472168, "num_tokens": 840458752.0, "step": 22025 }, { "epoch": 2.8019335962345755, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.456327438354492, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8807244300842285, "num_tokens": 840498441.0, "step": 22026 }, { "epoch": 2.8020608065131665, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.498208045959473, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8841291069984436, "num_tokens": 840537838.0, "step": 22027 }, { "epoch": 2.8021880167917566, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.557300567626953, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8635788559913635, "num_tokens": 840571770.0, "step": 22028 }, { "epoch": 2.8023152270703475, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.503963470458984, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8687112927436829, "num_tokens": 840607559.0, "step": 22029 }, { "epoch": 2.8024424373489376, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.42973518371582, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8602191805839539, "num_tokens": 840650156.0, "step": 22030 }, { "epoch": 2.8025696476275286, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.50029182434082, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8654988408088684, "num_tokens": 840686590.0, "step": 22031 }, { "epoch": 2.8026968579061187, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.539880752563477, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8666269779205322, "num_tokens": 840724445.0, "step": 22032 }, { "epoch": 2.802824068184709, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.53161334991455, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8887927532196045, "num_tokens": 840765047.0, "step": 22033 }, { "epoch": 2.8029512784632997, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.552861213684082, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.862243115901947, "num_tokens": 840802375.0, "step": 22034 }, { "epoch": 2.8030784887418903, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.535432815551758, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8730117678642273, "num_tokens": 840839994.0, "step": 22035 }, { "epoch": 2.803205699020481, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.526253700256348, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8727336525917053, "num_tokens": 840877144.0, "step": 22036 }, { "epoch": 2.8033329092990713, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.529547691345215, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8645477294921875, "num_tokens": 840918478.0, "step": 22037 }, { "epoch": 2.803460119577662, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.48110294342041, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8832679986953735, "num_tokens": 840958484.0, "step": 22038 }, { "epoch": 2.8035873298562524, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.47848892211914, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8763872385025024, "num_tokens": 841000035.0, "step": 22039 }, { "epoch": 2.803714540134843, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.50761890411377, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.868690550327301, "num_tokens": 841035836.0, "step": 22040 }, { "epoch": 2.8038417504134334, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.44472599029541, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8760038018226624, "num_tokens": 841069295.0, "step": 22041 }, { "epoch": 2.803968960692024, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.48857307434082, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8695183992385864, "num_tokens": 841108341.0, "step": 22042 }, { "epoch": 2.8040961709706145, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.543238639831543, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8839325308799744, "num_tokens": 841142945.0, "step": 22043 }, { "epoch": 2.804223381249205, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.486318588256836, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.886555016040802, "num_tokens": 841174137.0, "step": 22044 }, { "epoch": 2.8043505915277955, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.483918190002441, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8843938112258911, "num_tokens": 841215914.0, "step": 22045 }, { "epoch": 2.804477801806386, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.525848388671875, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8755916953086853, "num_tokens": 841256097.0, "step": 22046 }, { "epoch": 2.8046050120849766, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.47970199584961, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8783317804336548, "num_tokens": 841288997.0, "step": 22047 }, { "epoch": 2.804732222363567, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.592087745666504, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8697483539581299, "num_tokens": 841328053.0, "step": 22048 }, { "epoch": 2.8048594326421576, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.523035049438477, "learning_rate": 1e-06, "loss": 0.3466, "mean_token_accuracy": 0.8921303749084473, "num_tokens": 841366848.0, "step": 22049 }, { "epoch": 2.804986642920748, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.49365234375, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8827323317527771, "num_tokens": 841398568.0, "step": 22050 }, { "epoch": 2.8051138531993383, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.59029483795166, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.872802734375, "num_tokens": 841435722.0, "step": 22051 }, { "epoch": 2.8052410634779292, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.514322280883789, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8702422380447388, "num_tokens": 841473635.0, "step": 22052 }, { "epoch": 2.8053682737565193, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.495960235595703, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.876675009727478, "num_tokens": 841512383.0, "step": 22053 }, { "epoch": 2.8054954840351103, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.531968116760254, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.869861900806427, "num_tokens": 841543174.0, "step": 22054 }, { "epoch": 2.8056226943137004, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.501601219177246, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8791693449020386, "num_tokens": 841582185.0, "step": 22055 }, { "epoch": 2.805749904592291, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.528111457824707, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8690879344940186, "num_tokens": 841623179.0, "step": 22056 }, { "epoch": 2.8058771148708814, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.52833080291748, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8826161623001099, "num_tokens": 841661062.0, "step": 22057 }, { "epoch": 2.806004325149472, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.47596549987793, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8724482655525208, "num_tokens": 841704259.0, "step": 22058 }, { "epoch": 2.8061315354280625, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.54856014251709, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8723384141921997, "num_tokens": 841748396.0, "step": 22059 }, { "epoch": 2.806258745706653, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.464499473571777, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.892690122127533, "num_tokens": 841787809.0, "step": 22060 }, { "epoch": 2.8063859559852435, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.506219863891602, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8746742010116577, "num_tokens": 841824046.0, "step": 22061 }, { "epoch": 2.806513166263834, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.564820289611816, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8840069770812988, "num_tokens": 841856949.0, "step": 22062 }, { "epoch": 2.8066403765424246, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.497142791748047, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8603981733322144, "num_tokens": 841898644.0, "step": 22063 }, { "epoch": 2.806767586821015, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.470502853393555, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8765839338302612, "num_tokens": 841934795.0, "step": 22064 }, { "epoch": 2.8068947970996057, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.498871803283691, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8597671389579773, "num_tokens": 841976878.0, "step": 22065 }, { "epoch": 2.807022007378196, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.457844734191895, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8750462532043457, "num_tokens": 842010998.0, "step": 22066 }, { "epoch": 2.8071492176567867, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.50565242767334, "learning_rate": 1e-06, "loss": 0.322, "mean_token_accuracy": 0.8992040157318115, "num_tokens": 842051410.0, "step": 22067 }, { "epoch": 2.8072764279353772, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.412073135375977, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8770015835762024, "num_tokens": 842087954.0, "step": 22068 }, { "epoch": 2.8074036382139678, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.586610794067383, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8716001510620117, "num_tokens": 842128326.0, "step": 22069 }, { "epoch": 2.8075308484925583, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.508967399597168, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8543335199356079, "num_tokens": 842171143.0, "step": 22070 }, { "epoch": 2.807658058771149, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55942153930664, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8627686500549316, "num_tokens": 842212167.0, "step": 22071 }, { "epoch": 2.8077852690497394, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.463335037231445, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8655866980552673, "num_tokens": 842252576.0, "step": 22072 }, { "epoch": 2.80791247932833, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.657371520996094, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8709142208099365, "num_tokens": 842287812.0, "step": 22073 }, { "epoch": 2.80803968960692, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.503802299499512, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8705065846443176, "num_tokens": 842328289.0, "step": 22074 }, { "epoch": 2.808166899885511, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.535252571105957, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8778992891311646, "num_tokens": 842359182.0, "step": 22075 }, { "epoch": 2.808294110164101, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.504619598388672, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8645603060722351, "num_tokens": 842399820.0, "step": 22076 }, { "epoch": 2.808421320442692, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.448640823364258, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8730227947235107, "num_tokens": 842437852.0, "step": 22077 }, { "epoch": 2.808548530721282, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.553994178771973, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8614538908004761, "num_tokens": 842476729.0, "step": 22078 }, { "epoch": 2.808675740999873, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.41247272491455, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8622422218322754, "num_tokens": 842514590.0, "step": 22079 }, { "epoch": 2.808802951278463, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.45771598815918, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8726378679275513, "num_tokens": 842556502.0, "step": 22080 }, { "epoch": 2.8089301615570537, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.44111442565918, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8630327582359314, "num_tokens": 842594299.0, "step": 22081 }, { "epoch": 2.809057371835644, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.541796684265137, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8713909387588501, "num_tokens": 842633189.0, "step": 22082 }, { "epoch": 2.8091845821142347, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.425250053405762, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8765815496444702, "num_tokens": 842673626.0, "step": 22083 }, { "epoch": 2.8093117923928252, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.488140106201172, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8666506409645081, "num_tokens": 842712574.0, "step": 22084 }, { "epoch": 2.8094390026714158, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.502395629882812, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8716868162155151, "num_tokens": 842749534.0, "step": 22085 }, { "epoch": 2.8095662129500063, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.53951644897461, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.865127682685852, "num_tokens": 842784989.0, "step": 22086 }, { "epoch": 2.809693423228597, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.488208770751953, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8747384548187256, "num_tokens": 842823858.0, "step": 22087 }, { "epoch": 2.8098206335071874, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.491482734680176, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8753149509429932, "num_tokens": 842858849.0, "step": 22088 }, { "epoch": 2.809947843785778, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.451361656188965, "learning_rate": 1e-06, "loss": 0.3399, "mean_token_accuracy": 0.8935869336128235, "num_tokens": 842895809.0, "step": 22089 }, { "epoch": 2.8100750540643684, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.488606452941895, "learning_rate": 1e-06, "loss": 0.3159, "mean_token_accuracy": 0.900943398475647, "num_tokens": 842933650.0, "step": 22090 }, { "epoch": 2.810202264342959, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.543437957763672, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.879718005657196, "num_tokens": 842967200.0, "step": 22091 }, { "epoch": 2.8103294746215495, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.424838066101074, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8655941486358643, "num_tokens": 843002498.0, "step": 22092 }, { "epoch": 2.81045668490014, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.508749008178711, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8791471719741821, "num_tokens": 843042627.0, "step": 22093 }, { "epoch": 2.8105838951787305, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.465580940246582, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8715260624885559, "num_tokens": 843076862.0, "step": 22094 }, { "epoch": 2.810711105457321, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.524673461914062, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8737539052963257, "num_tokens": 843115000.0, "step": 22095 }, { "epoch": 2.8108383157359116, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.462455749511719, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8808602690696716, "num_tokens": 843152408.0, "step": 22096 }, { "epoch": 2.810965526014502, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.475464820861816, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8755383491516113, "num_tokens": 843189269.0, "step": 22097 }, { "epoch": 2.8110927362930926, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.519477844238281, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8697428703308105, "num_tokens": 843227148.0, "step": 22098 }, { "epoch": 2.8112199465716827, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.448521614074707, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8652132749557495, "num_tokens": 843259736.0, "step": 22099 }, { "epoch": 2.8113471568502737, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.427669525146484, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8702982068061829, "num_tokens": 843302598.0, "step": 22100 }, { "epoch": 2.811474367128864, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.566819190979004, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8897321224212646, "num_tokens": 843337615.0, "step": 22101 }, { "epoch": 2.8116015774074548, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.509908676147461, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8888672590255737, "num_tokens": 843378845.0, "step": 22102 }, { "epoch": 2.811728787686045, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.58561897277832, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8655444979667664, "num_tokens": 843414128.0, "step": 22103 }, { "epoch": 2.811855997964636, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.449385643005371, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8893285989761353, "num_tokens": 843454255.0, "step": 22104 }, { "epoch": 2.811983208243226, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.534330368041992, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8755109310150146, "num_tokens": 843498332.0, "step": 22105 }, { "epoch": 2.8121104185218164, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.498769760131836, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8643811941146851, "num_tokens": 843539196.0, "step": 22106 }, { "epoch": 2.812237628800407, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.537148475646973, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8777166604995728, "num_tokens": 843573767.0, "step": 22107 }, { "epoch": 2.8123648390789975, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.532962799072266, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8621523380279541, "num_tokens": 843611156.0, "step": 22108 }, { "epoch": 2.812492049357588, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.503730773925781, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8798846006393433, "num_tokens": 843644026.0, "step": 22109 }, { "epoch": 2.8126192596361785, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.559104919433594, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8775786757469177, "num_tokens": 843683477.0, "step": 22110 }, { "epoch": 2.812746469914769, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.533895492553711, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8620442152023315, "num_tokens": 843722815.0, "step": 22111 }, { "epoch": 2.8128736801933596, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.509626388549805, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8754285573959351, "num_tokens": 843763843.0, "step": 22112 }, { "epoch": 2.81300089047195, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.487861633300781, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8664975762367249, "num_tokens": 843805327.0, "step": 22113 }, { "epoch": 2.8131281007505406, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.460331916809082, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8747864961624146, "num_tokens": 843850132.0, "step": 22114 }, { "epoch": 2.813255311029131, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.530715942382812, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8744859099388123, "num_tokens": 843887425.0, "step": 22115 }, { "epoch": 2.8133825213077217, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.573198318481445, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8830463886260986, "num_tokens": 843920728.0, "step": 22116 }, { "epoch": 2.8135097315863122, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.486952781677246, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8733033537864685, "num_tokens": 843959733.0, "step": 22117 }, { "epoch": 2.8136369418649028, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.574882507324219, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8756566047668457, "num_tokens": 844005616.0, "step": 22118 }, { "epoch": 2.8137641521434933, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.491952896118164, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.8879538774490356, "num_tokens": 844034216.0, "step": 22119 }, { "epoch": 2.813891362422084, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.482824325561523, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8708846569061279, "num_tokens": 844071173.0, "step": 22120 }, { "epoch": 2.8140185727006743, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.492416381835938, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.872484564781189, "num_tokens": 844113665.0, "step": 22121 }, { "epoch": 2.814145782979265, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.510359764099121, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8753088712692261, "num_tokens": 844154063.0, "step": 22122 }, { "epoch": 2.8142729932578554, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.51503849029541, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8932102918624878, "num_tokens": 844185634.0, "step": 22123 }, { "epoch": 2.8144002035364455, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.473037719726562, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8649930357933044, "num_tokens": 844221755.0, "step": 22124 }, { "epoch": 2.8145274138150365, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.532320976257324, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8792392015457153, "num_tokens": 844256230.0, "step": 22125 }, { "epoch": 2.8146546240936265, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.504460334777832, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8685916662216187, "num_tokens": 844292823.0, "step": 22126 }, { "epoch": 2.8147818343722175, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.53386402130127, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8525636196136475, "num_tokens": 844331843.0, "step": 22127 }, { "epoch": 2.8149090446508076, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.504463195800781, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8844404220581055, "num_tokens": 844375823.0, "step": 22128 }, { "epoch": 2.815036254929398, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.488548278808594, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8744038343429565, "num_tokens": 844415187.0, "step": 22129 }, { "epoch": 2.8151634652079887, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.506342887878418, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8778548240661621, "num_tokens": 844454986.0, "step": 22130 }, { "epoch": 2.815290675486579, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.51590347290039, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8805079460144043, "num_tokens": 844495514.0, "step": 22131 }, { "epoch": 2.8154178857651697, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.472947120666504, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8738194704055786, "num_tokens": 844532406.0, "step": 22132 }, { "epoch": 2.8155450960437602, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.542740821838379, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8737431764602661, "num_tokens": 844569561.0, "step": 22133 }, { "epoch": 2.8156723063223508, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.500805854797363, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8792930841445923, "num_tokens": 844613504.0, "step": 22134 }, { "epoch": 2.8157995166009413, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.553754806518555, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8722163438796997, "num_tokens": 844649255.0, "step": 22135 }, { "epoch": 2.815926726879532, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.598604202270508, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8766239881515503, "num_tokens": 844687202.0, "step": 22136 }, { "epoch": 2.8160539371581224, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.552722930908203, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.87558913230896, "num_tokens": 844723096.0, "step": 22137 }, { "epoch": 2.816181147436713, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.578283309936523, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8763698935508728, "num_tokens": 844759754.0, "step": 22138 }, { "epoch": 2.8163083577153034, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.572915077209473, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.878257155418396, "num_tokens": 844798372.0, "step": 22139 }, { "epoch": 2.816435567993894, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.572383880615234, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8687552213668823, "num_tokens": 844832738.0, "step": 22140 }, { "epoch": 2.8165627782724845, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.560722351074219, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8752725124359131, "num_tokens": 844871684.0, "step": 22141 }, { "epoch": 2.816689988551075, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.52098560333252, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.86753910779953, "num_tokens": 844912140.0, "step": 22142 }, { "epoch": 2.8168171988296655, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.560821533203125, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8623532056808472, "num_tokens": 844948256.0, "step": 22143 }, { "epoch": 2.816944409108256, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.49071979522705, "learning_rate": 1e-06, "loss": 0.3459, "mean_token_accuracy": 0.8901281356811523, "num_tokens": 844983821.0, "step": 22144 }, { "epoch": 2.8170716193868466, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.500935554504395, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8786616921424866, "num_tokens": 845020622.0, "step": 22145 }, { "epoch": 2.817198829665437, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.529415130615234, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8654205799102783, "num_tokens": 845056748.0, "step": 22146 }, { "epoch": 2.8173260399440276, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.47825813293457, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8564265966415405, "num_tokens": 845096294.0, "step": 22147 }, { "epoch": 2.817453250222618, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.966681480407715, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8893541097640991, "num_tokens": 845135132.0, "step": 22148 }, { "epoch": 2.8175804605012083, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.493454933166504, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8828543424606323, "num_tokens": 845172164.0, "step": 22149 }, { "epoch": 2.8177076707797992, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.398974418640137, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8755409717559814, "num_tokens": 845210270.0, "step": 22150 }, { "epoch": 2.8178348810583893, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.556798934936523, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.878364086151123, "num_tokens": 845251863.0, "step": 22151 }, { "epoch": 2.8179620913369803, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.450407028198242, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8782600164413452, "num_tokens": 845290262.0, "step": 22152 }, { "epoch": 2.8180893016155704, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.478625297546387, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8821761608123779, "num_tokens": 845326894.0, "step": 22153 }, { "epoch": 2.818216511894161, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.422065734863281, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.883269727230072, "num_tokens": 845367545.0, "step": 22154 }, { "epoch": 2.8183437221727514, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.511655807495117, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.860418438911438, "num_tokens": 845400920.0, "step": 22155 }, { "epoch": 2.818470932451342, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.484777450561523, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8715716004371643, "num_tokens": 845436279.0, "step": 22156 }, { "epoch": 2.8185981427299325, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.480511665344238, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.878313422203064, "num_tokens": 845476308.0, "step": 22157 }, { "epoch": 2.818725353008523, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.45173454284668, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8713314533233643, "num_tokens": 845516624.0, "step": 22158 }, { "epoch": 2.8188525632871135, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.462383270263672, "learning_rate": 1e-06, "loss": 0.3398, "mean_token_accuracy": 0.8921290040016174, "num_tokens": 845552385.0, "step": 22159 }, { "epoch": 2.818979773565704, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.484620094299316, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.883927047252655, "num_tokens": 845592600.0, "step": 22160 }, { "epoch": 2.8191069838442946, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.503737449645996, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.882419228553772, "num_tokens": 845630028.0, "step": 22161 }, { "epoch": 2.819234194122885, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.468547821044922, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8707988858222961, "num_tokens": 845665874.0, "step": 22162 }, { "epoch": 2.8193614044014756, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.50875186920166, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8633841276168823, "num_tokens": 845702738.0, "step": 22163 }, { "epoch": 2.819488614680066, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.506352424621582, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8783025741577148, "num_tokens": 845744785.0, "step": 22164 }, { "epoch": 2.8196158249586567, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.527588844299316, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8748512268066406, "num_tokens": 845779196.0, "step": 22165 }, { "epoch": 2.8197430352372472, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.459097862243652, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8651726245880127, "num_tokens": 845819016.0, "step": 22166 }, { "epoch": 2.8198702455158378, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.569866180419922, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8769257664680481, "num_tokens": 845855014.0, "step": 22167 }, { "epoch": 2.8199974557944283, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.470768928527832, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8829801678657532, "num_tokens": 845895491.0, "step": 22168 }, { "epoch": 2.820124666073019, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.534370422363281, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8657407760620117, "num_tokens": 845931560.0, "step": 22169 }, { "epoch": 2.8202518763516093, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.489352226257324, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.874550998210907, "num_tokens": 845966701.0, "step": 22170 }, { "epoch": 2.8203790866302, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.484936714172363, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8919644355773926, "num_tokens": 846002327.0, "step": 22171 }, { "epoch": 2.82050629690879, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.502366065979004, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8755261898040771, "num_tokens": 846042157.0, "step": 22172 }, { "epoch": 2.820633507187381, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.462859153747559, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8592365980148315, "num_tokens": 846082851.0, "step": 22173 }, { "epoch": 2.820760717465971, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.558582305908203, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8736749887466431, "num_tokens": 846116846.0, "step": 22174 }, { "epoch": 2.820887927744562, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.544609069824219, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8773133754730225, "num_tokens": 846154373.0, "step": 22175 }, { "epoch": 2.821015138023152, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.492042541503906, "learning_rate": 1e-06, "loss": 0.337, "mean_token_accuracy": 0.8920828104019165, "num_tokens": 846195329.0, "step": 22176 }, { "epoch": 2.821142348301743, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.540212631225586, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8863469362258911, "num_tokens": 846233078.0, "step": 22177 }, { "epoch": 2.821269558580333, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.535991668701172, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8737646341323853, "num_tokens": 846269030.0, "step": 22178 }, { "epoch": 2.8213967688589237, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.509230613708496, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8707708120346069, "num_tokens": 846307269.0, "step": 22179 }, { "epoch": 2.821523979137514, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.538296699523926, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8662174940109253, "num_tokens": 846339116.0, "step": 22180 }, { "epoch": 2.8216511894161047, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.427910804748535, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8665708899497986, "num_tokens": 846384968.0, "step": 22181 }, { "epoch": 2.8217783996946952, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.525385856628418, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8593740463256836, "num_tokens": 846426235.0, "step": 22182 }, { "epoch": 2.8219056099732858, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.543647766113281, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8617256283760071, "num_tokens": 846465715.0, "step": 22183 }, { "epoch": 2.8220328202518763, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.469162940979004, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.861562967300415, "num_tokens": 846499972.0, "step": 22184 }, { "epoch": 2.822160030530467, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.502144813537598, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8729420900344849, "num_tokens": 846535906.0, "step": 22185 }, { "epoch": 2.8222872408090574, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.516327857971191, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8711204528808594, "num_tokens": 846572723.0, "step": 22186 }, { "epoch": 2.822414451087648, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.547274589538574, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8791482448577881, "num_tokens": 846613501.0, "step": 22187 }, { "epoch": 2.8225416613662384, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.507942199707031, "learning_rate": 1e-06, "loss": 0.3411, "mean_token_accuracy": 0.8945607542991638, "num_tokens": 846650001.0, "step": 22188 }, { "epoch": 2.822668871644829, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.479637145996094, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8633257150650024, "num_tokens": 846690636.0, "step": 22189 }, { "epoch": 2.8227960819234195, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.621074676513672, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.853938102722168, "num_tokens": 846728513.0, "step": 22190 }, { "epoch": 2.82292329220201, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.506513595581055, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8609879612922668, "num_tokens": 846771989.0, "step": 22191 }, { "epoch": 2.8230505024806005, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.590527534484863, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.8918130397796631, "num_tokens": 846808393.0, "step": 22192 }, { "epoch": 2.823177712759191, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.519599914550781, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8647148609161377, "num_tokens": 846846928.0, "step": 22193 }, { "epoch": 2.8233049230377816, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.579667091369629, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8648422360420227, "num_tokens": 846884637.0, "step": 22194 }, { "epoch": 2.823432133316372, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.626053810119629, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8769668340682983, "num_tokens": 846925749.0, "step": 22195 }, { "epoch": 2.8235593435949626, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.498984336853027, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8606168031692505, "num_tokens": 846967191.0, "step": 22196 }, { "epoch": 2.8236865538735527, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.601018905639648, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8793894052505493, "num_tokens": 847003280.0, "step": 22197 }, { "epoch": 2.8238137641521437, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.490782737731934, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8628355264663696, "num_tokens": 847042179.0, "step": 22198 }, { "epoch": 2.8239409744307338, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.622723579406738, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8704231977462769, "num_tokens": 847083008.0, "step": 22199 }, { "epoch": 2.8240681847093247, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.500138282775879, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8537349700927734, "num_tokens": 847124009.0, "step": 22200 }, { "epoch": 2.824195394987915, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.594832420349121, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8684046268463135, "num_tokens": 847163526.0, "step": 22201 }, { "epoch": 2.824322605266506, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.616252899169922, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8722743988037109, "num_tokens": 847201374.0, "step": 22202 }, { "epoch": 2.824449815545096, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55769157409668, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8596282005310059, "num_tokens": 847239469.0, "step": 22203 }, { "epoch": 2.8245770258236864, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61038875579834, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8720242977142334, "num_tokens": 847275964.0, "step": 22204 }, { "epoch": 2.824704236102277, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.622426986694336, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8755805492401123, "num_tokens": 847316815.0, "step": 22205 }, { "epoch": 2.8248314463808675, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.552583694458008, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8797881603240967, "num_tokens": 847350949.0, "step": 22206 }, { "epoch": 2.824958656659458, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.574667930603027, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8785839676856995, "num_tokens": 847388945.0, "step": 22207 }, { "epoch": 2.8250858669380485, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.53325366973877, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.885337233543396, "num_tokens": 847435114.0, "step": 22208 }, { "epoch": 2.825213077216639, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.580826759338379, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8657270073890686, "num_tokens": 847470815.0, "step": 22209 }, { "epoch": 2.8253402874952296, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.6007719039917, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8715784549713135, "num_tokens": 847509379.0, "step": 22210 }, { "epoch": 2.82546749777382, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.591763496398926, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8774386644363403, "num_tokens": 847547427.0, "step": 22211 }, { "epoch": 2.8255947080524106, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.566554069519043, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8574633598327637, "num_tokens": 847584740.0, "step": 22212 }, { "epoch": 2.825721918331001, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.597078323364258, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8620749115943909, "num_tokens": 847626201.0, "step": 22213 }, { "epoch": 2.8258491286095917, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.536417961120605, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8881823420524597, "num_tokens": 847661311.0, "step": 22214 }, { "epoch": 2.8259763388881822, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61896800994873, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8723360300064087, "num_tokens": 847695234.0, "step": 22215 }, { "epoch": 2.8261035491667728, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.594826698303223, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8903088569641113, "num_tokens": 847738848.0, "step": 22216 }, { "epoch": 2.8262307594453633, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.60003662109375, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8773118853569031, "num_tokens": 847775952.0, "step": 22217 }, { "epoch": 2.826357969723954, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.604867935180664, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8895764946937561, "num_tokens": 847809381.0, "step": 22218 }, { "epoch": 2.8264851800025443, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.570829391479492, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.870876133441925, "num_tokens": 847849835.0, "step": 22219 }, { "epoch": 2.826612390281135, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.506503105163574, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8716449737548828, "num_tokens": 847882726.0, "step": 22220 }, { "epoch": 2.8267396005597254, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.637828826904297, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8732218146324158, "num_tokens": 847920663.0, "step": 22221 }, { "epoch": 2.8268668108383155, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.589807510375977, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8818777799606323, "num_tokens": 847962204.0, "step": 22222 }, { "epoch": 2.8269940211169065, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.519474983215332, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8669130802154541, "num_tokens": 848003722.0, "step": 22223 }, { "epoch": 2.8271212313954965, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55591869354248, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8823365569114685, "num_tokens": 848045335.0, "step": 22224 }, { "epoch": 2.8272484416740875, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.452335357666016, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8665831089019775, "num_tokens": 848088817.0, "step": 22225 }, { "epoch": 2.8273756519526776, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.576285362243652, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8648419380187988, "num_tokens": 848120993.0, "step": 22226 }, { "epoch": 2.827502862231268, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.53982162475586, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8669851422309875, "num_tokens": 848162542.0, "step": 22227 }, { "epoch": 2.8276300725098586, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55010986328125, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8873735666275024, "num_tokens": 848198194.0, "step": 22228 }, { "epoch": 2.827757282788449, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.538235664367676, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8688143491744995, "num_tokens": 848240575.0, "step": 22229 }, { "epoch": 2.8278844930670397, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.51440143585205, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8736201524734497, "num_tokens": 848289026.0, "step": 22230 }, { "epoch": 2.8280117033456302, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.632472038269043, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8760536313056946, "num_tokens": 848326438.0, "step": 22231 }, { "epoch": 2.8281389136242208, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.542960166931152, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8808397054672241, "num_tokens": 848368106.0, "step": 22232 }, { "epoch": 2.8282661239028113, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.542304039001465, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8707523345947266, "num_tokens": 848409775.0, "step": 22233 }, { "epoch": 2.828393334181402, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.573247909545898, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8742324113845825, "num_tokens": 848442149.0, "step": 22234 }, { "epoch": 2.8285205444599923, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.568799018859863, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.867997944355011, "num_tokens": 848475468.0, "step": 22235 }, { "epoch": 2.828647754738583, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.508218765258789, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8739892244338989, "num_tokens": 848509565.0, "step": 22236 }, { "epoch": 2.8287749650171734, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.550670623779297, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8756040334701538, "num_tokens": 848546357.0, "step": 22237 }, { "epoch": 2.828902175295764, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.515552520751953, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8632155656814575, "num_tokens": 848587661.0, "step": 22238 }, { "epoch": 2.8290293855743545, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.445380210876465, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.877842366695404, "num_tokens": 848628250.0, "step": 22239 }, { "epoch": 2.829156595852945, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.581031799316406, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8739440441131592, "num_tokens": 848670452.0, "step": 22240 }, { "epoch": 2.8292838061315355, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.614081382751465, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8759711384773254, "num_tokens": 848703592.0, "step": 22241 }, { "epoch": 2.829411016410126, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.53997802734375, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8580424785614014, "num_tokens": 848739291.0, "step": 22242 }, { "epoch": 2.8295382266887166, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.543437004089355, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8685383796691895, "num_tokens": 848776039.0, "step": 22243 }, { "epoch": 2.829665436967307, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.56446361541748, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8765438795089722, "num_tokens": 848814799.0, "step": 22244 }, { "epoch": 2.8297926472458976, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.525389671325684, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8751214742660522, "num_tokens": 848860123.0, "step": 22245 }, { "epoch": 2.829919857524488, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.498871803283691, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8905245661735535, "num_tokens": 848900447.0, "step": 22246 }, { "epoch": 2.8300470678030782, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.503986358642578, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8681267499923706, "num_tokens": 848945385.0, "step": 22247 }, { "epoch": 2.830174278081669, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.56107234954834, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8638931512832642, "num_tokens": 848978783.0, "step": 22248 }, { "epoch": 2.8303014883602593, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.555715560913086, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8629481792449951, "num_tokens": 849017324.0, "step": 22249 }, { "epoch": 2.8304286986388503, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.58221435546875, "learning_rate": 1e-06, "loss": 0.3435, "mean_token_accuracy": 0.8939054608345032, "num_tokens": 849045928.0, "step": 22250 }, { "epoch": 2.8305559089174404, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.510433197021484, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8739705085754395, "num_tokens": 849080855.0, "step": 22251 }, { "epoch": 2.830683119196031, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.620908737182617, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8773267269134521, "num_tokens": 849114835.0, "step": 22252 }, { "epoch": 2.8308103294746214, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.484026908874512, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8629863262176514, "num_tokens": 849152632.0, "step": 22253 }, { "epoch": 2.830937539753212, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.535530090332031, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8832449316978455, "num_tokens": 849189911.0, "step": 22254 }, { "epoch": 2.8310647500318025, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.532110214233398, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.873420774936676, "num_tokens": 849227847.0, "step": 22255 }, { "epoch": 2.831191960310393, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.56309700012207, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8793483972549438, "num_tokens": 849265778.0, "step": 22256 }, { "epoch": 2.8313191705889835, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.510891914367676, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8738266229629517, "num_tokens": 849302488.0, "step": 22257 }, { "epoch": 2.831446380867574, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.555325508117676, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8886837363243103, "num_tokens": 849337483.0, "step": 22258 }, { "epoch": 2.8315735911461646, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.598694801330566, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8802422285079956, "num_tokens": 849378491.0, "step": 22259 }, { "epoch": 2.831700801424755, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.478570938110352, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8812639713287354, "num_tokens": 849418897.0, "step": 22260 }, { "epoch": 2.8318280117033456, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.515643119812012, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8787543773651123, "num_tokens": 849457057.0, "step": 22261 }, { "epoch": 2.831955221981936, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.53044605255127, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8787760734558105, "num_tokens": 849488825.0, "step": 22262 }, { "epoch": 2.8320824322605267, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.517581939697266, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8810368776321411, "num_tokens": 849528336.0, "step": 22263 }, { "epoch": 2.8322096425391172, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.544390678405762, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8872898817062378, "num_tokens": 849565373.0, "step": 22264 }, { "epoch": 2.8323368528177078, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.512825012207031, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8848679661750793, "num_tokens": 849600909.0, "step": 22265 }, { "epoch": 2.8324640630962983, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.60549259185791, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8654109835624695, "num_tokens": 849639499.0, "step": 22266 }, { "epoch": 2.832591273374889, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.493254661560059, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.885261595249176, "num_tokens": 849677232.0, "step": 22267 }, { "epoch": 2.8327184836534793, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.626506805419922, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8806747794151306, "num_tokens": 849717187.0, "step": 22268 }, { "epoch": 2.83284569393207, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.509571075439453, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8844032287597656, "num_tokens": 849756545.0, "step": 22269 }, { "epoch": 2.83297290421066, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.593897819519043, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8786553740501404, "num_tokens": 849788670.0, "step": 22270 }, { "epoch": 2.833100114489251, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.548081398010254, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.874022364616394, "num_tokens": 849825715.0, "step": 22271 }, { "epoch": 2.833227324767841, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.550683975219727, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.88070148229599, "num_tokens": 849856435.0, "step": 22272 }, { "epoch": 2.833354535046432, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.596988677978516, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8889824151992798, "num_tokens": 849895634.0, "step": 22273 }, { "epoch": 2.833481745325022, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.485797882080078, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8539674282073975, "num_tokens": 849938646.0, "step": 22274 }, { "epoch": 2.833608955603613, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.585368156433105, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8711775541305542, "num_tokens": 849978642.0, "step": 22275 }, { "epoch": 2.833736165882203, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.502652168273926, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8743773102760315, "num_tokens": 850017820.0, "step": 22276 }, { "epoch": 2.8338633761607936, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.564952850341797, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8555940985679626, "num_tokens": 850056619.0, "step": 22277 }, { "epoch": 2.833990586439384, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.610150337219238, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8734734058380127, "num_tokens": 850092456.0, "step": 22278 }, { "epoch": 2.8341177967179747, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.51638126373291, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8711000680923462, "num_tokens": 850131598.0, "step": 22279 }, { "epoch": 2.8342450069965652, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.622485160827637, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8649044036865234, "num_tokens": 850170884.0, "step": 22280 }, { "epoch": 2.8343722172751558, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.508090019226074, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8739526867866516, "num_tokens": 850206989.0, "step": 22281 }, { "epoch": 2.8344994275537463, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.541428565979004, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.870802640914917, "num_tokens": 850248602.0, "step": 22282 }, { "epoch": 2.834626637832337, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.652801513671875, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8671819567680359, "num_tokens": 850289790.0, "step": 22283 }, { "epoch": 2.8347538481109273, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.508289337158203, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8826780319213867, "num_tokens": 850326688.0, "step": 22284 }, { "epoch": 2.834881058389518, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.573773384094238, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8855346441268921, "num_tokens": 850366460.0, "step": 22285 }, { "epoch": 2.8350082686681084, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.572590827941895, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8804534673690796, "num_tokens": 850400046.0, "step": 22286 }, { "epoch": 2.835135478946699, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.618075370788574, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8661131858825684, "num_tokens": 850434374.0, "step": 22287 }, { "epoch": 2.8352626892252895, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.577058792114258, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8763088583946228, "num_tokens": 850469056.0, "step": 22288 }, { "epoch": 2.83538989950388, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55489730834961, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8714768290519714, "num_tokens": 850505805.0, "step": 22289 }, { "epoch": 2.8355171097824705, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.603342056274414, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8626598119735718, "num_tokens": 850550056.0, "step": 22290 }, { "epoch": 2.835644320061061, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.549829483032227, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8817657232284546, "num_tokens": 850585075.0, "step": 22291 }, { "epoch": 2.8357715303396516, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.561034202575684, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8840370178222656, "num_tokens": 850625758.0, "step": 22292 }, { "epoch": 2.835898740618242, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.524999618530273, "learning_rate": 1e-06, "loss": 0.3486, "mean_token_accuracy": 0.8870165348052979, "num_tokens": 850664608.0, "step": 22293 }, { "epoch": 2.8360259508968326, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.560026168823242, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8817451000213623, "num_tokens": 850705897.0, "step": 22294 }, { "epoch": 2.8361531611754227, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.528204917907715, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8698228597640991, "num_tokens": 850746195.0, "step": 22295 }, { "epoch": 2.8362803714540137, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.568048477172852, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8820896148681641, "num_tokens": 850785183.0, "step": 22296 }, { "epoch": 2.8364075817326038, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.554243087768555, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8798283338546753, "num_tokens": 850819558.0, "step": 22297 }, { "epoch": 2.8365347920111947, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.535710334777832, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8656479120254517, "num_tokens": 850857385.0, "step": 22298 }, { "epoch": 2.836662002289785, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.624588966369629, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8694416284561157, "num_tokens": 850895297.0, "step": 22299 }, { "epoch": 2.836789212568376, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.535955429077148, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.874373733997345, "num_tokens": 850930982.0, "step": 22300 }, { "epoch": 2.836916422846966, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.581543922424316, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8703649044036865, "num_tokens": 850969032.0, "step": 22301 }, { "epoch": 2.8370436331255564, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.551948547363281, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8776189088821411, "num_tokens": 851008325.0, "step": 22302 }, { "epoch": 2.837170843404147, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61815071105957, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8761791586875916, "num_tokens": 851044289.0, "step": 22303 }, { "epoch": 2.8372980536827375, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.57055950164795, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8814548254013062, "num_tokens": 851079887.0, "step": 22304 }, { "epoch": 2.837425263961328, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.550148963928223, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8772363066673279, "num_tokens": 851124535.0, "step": 22305 }, { "epoch": 2.8375524742399185, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.630636215209961, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8829958438873291, "num_tokens": 851162982.0, "step": 22306 }, { "epoch": 2.837679684518509, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.540836334228516, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8817684054374695, "num_tokens": 851201864.0, "step": 22307 }, { "epoch": 2.8378068947970996, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.553277969360352, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8782910704612732, "num_tokens": 851241659.0, "step": 22308 }, { "epoch": 2.83793410507569, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.551704406738281, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8669185042381287, "num_tokens": 851278861.0, "step": 22309 }, { "epoch": 2.8380613153542806, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.60746955871582, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8759260177612305, "num_tokens": 851314442.0, "step": 22310 }, { "epoch": 2.838188525632871, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.608170509338379, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8702079057693481, "num_tokens": 851356322.0, "step": 22311 }, { "epoch": 2.8383157359114617, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.59038257598877, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8631687760353088, "num_tokens": 851395374.0, "step": 22312 }, { "epoch": 2.838442946190052, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.597016334533691, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.87267005443573, "num_tokens": 851430957.0, "step": 22313 }, { "epoch": 2.8385701564686427, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.481682777404785, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8692402839660645, "num_tokens": 851474357.0, "step": 22314 }, { "epoch": 2.8386973667472333, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.552663803100586, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8870365619659424, "num_tokens": 851510346.0, "step": 22315 }, { "epoch": 2.838824577025824, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.515891075134277, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8766933083534241, "num_tokens": 851546798.0, "step": 22316 }, { "epoch": 2.8389517873044143, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.61733341217041, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8797056674957275, "num_tokens": 851587819.0, "step": 22317 }, { "epoch": 2.839078997583005, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.573884010314941, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8793805837631226, "num_tokens": 851628764.0, "step": 22318 }, { "epoch": 2.8392062078615954, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.526391983032227, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8611457347869873, "num_tokens": 851672206.0, "step": 22319 }, { "epoch": 2.8393334181401855, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.537015914916992, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8694444894790649, "num_tokens": 851709969.0, "step": 22320 }, { "epoch": 2.8394606284187764, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.557820320129395, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.868507444858551, "num_tokens": 851749015.0, "step": 22321 }, { "epoch": 2.8395878386973665, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.545187950134277, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8784992098808289, "num_tokens": 851787210.0, "step": 22322 }, { "epoch": 2.8397150489759575, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.561845779418945, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8682276010513306, "num_tokens": 851824677.0, "step": 22323 }, { "epoch": 2.8398422592545476, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.596848487854004, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8794502019882202, "num_tokens": 851863614.0, "step": 22324 }, { "epoch": 2.839969469533138, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.573731422424316, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8681405186653137, "num_tokens": 851908153.0, "step": 22325 }, { "epoch": 2.8400966798117286, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.537049293518066, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8709964156150818, "num_tokens": 851946473.0, "step": 22326 }, { "epoch": 2.840223890090319, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61653995513916, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8638322353363037, "num_tokens": 851991600.0, "step": 22327 }, { "epoch": 2.8403511003689097, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61800479888916, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8765088319778442, "num_tokens": 852029468.0, "step": 22328 }, { "epoch": 2.8404783106475002, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.550206184387207, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8811604380607605, "num_tokens": 852061400.0, "step": 22329 }, { "epoch": 2.8406055209260908, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.536383628845215, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8735756874084473, "num_tokens": 852100596.0, "step": 22330 }, { "epoch": 2.8407327312046813, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.581642150878906, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8773137331008911, "num_tokens": 852140320.0, "step": 22331 }, { "epoch": 2.840859941483272, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.63699722290039, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8733923435211182, "num_tokens": 852175969.0, "step": 22332 }, { "epoch": 2.8409871517618623, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.532713890075684, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.861598551273346, "num_tokens": 852210816.0, "step": 22333 }, { "epoch": 2.841114362040453, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.587827682495117, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8738112449645996, "num_tokens": 852246684.0, "step": 22334 }, { "epoch": 2.8412415723190434, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.555403709411621, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.877646803855896, "num_tokens": 852285283.0, "step": 22335 }, { "epoch": 2.841368782597634, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.603255271911621, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8820862770080566, "num_tokens": 852323310.0, "step": 22336 }, { "epoch": 2.8414959928762245, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.591520309448242, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8782247304916382, "num_tokens": 852360850.0, "step": 22337 }, { "epoch": 2.841623203154815, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.638127326965332, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8682769536972046, "num_tokens": 852399320.0, "step": 22338 }, { "epoch": 2.8417504134334055, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.565604209899902, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8725658655166626, "num_tokens": 852430980.0, "step": 22339 }, { "epoch": 2.841877623711996, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.67193603515625, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8856978416442871, "num_tokens": 852465688.0, "step": 22340 }, { "epoch": 2.8420048339905866, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.584939002990723, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.872937023639679, "num_tokens": 852504372.0, "step": 22341 }, { "epoch": 2.842132044269177, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.651528358459473, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8792202472686768, "num_tokens": 852542232.0, "step": 22342 }, { "epoch": 2.8422592545477676, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.563879013061523, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8653264045715332, "num_tokens": 852580005.0, "step": 22343 }, { "epoch": 2.842386464826358, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.60986042022705, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8859032392501831, "num_tokens": 852619912.0, "step": 22344 }, { "epoch": 2.8425136751049482, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.626740455627441, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8773866295814514, "num_tokens": 852654349.0, "step": 22345 }, { "epoch": 2.842640885383539, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.600522994995117, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8672173023223877, "num_tokens": 852691300.0, "step": 22346 }, { "epoch": 2.8427680956621293, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.619028091430664, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8730462789535522, "num_tokens": 852731228.0, "step": 22347 }, { "epoch": 2.8428953059407203, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.569132804870605, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8738191723823547, "num_tokens": 852768333.0, "step": 22348 }, { "epoch": 2.8430225162193103, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.668923377990723, "learning_rate": 1e-06, "loss": 0.4948, "mean_token_accuracy": 0.8432374000549316, "num_tokens": 852810962.0, "step": 22349 }, { "epoch": 2.843149726497901, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.581254005432129, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8916476964950562, "num_tokens": 852844326.0, "step": 22350 }, { "epoch": 2.8432769367764914, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.593743324279785, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8696179389953613, "num_tokens": 852877737.0, "step": 22351 }, { "epoch": 2.843404147055082, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.607824325561523, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8812962770462036, "num_tokens": 852922085.0, "step": 22352 }, { "epoch": 2.8435313573336725, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.608800888061523, "learning_rate": 1e-06, "loss": 0.3421, "mean_token_accuracy": 0.8910359740257263, "num_tokens": 852960742.0, "step": 22353 }, { "epoch": 2.843658567612263, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.544994354248047, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8789785504341125, "num_tokens": 852993502.0, "step": 22354 }, { "epoch": 2.8437857778908535, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.594980239868164, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8788334131240845, "num_tokens": 853036844.0, "step": 22355 }, { "epoch": 2.843912988169444, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.581445693969727, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8739668130874634, "num_tokens": 853070851.0, "step": 22356 }, { "epoch": 2.8440401984480346, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.579051971435547, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8745709657669067, "num_tokens": 853106116.0, "step": 22357 }, { "epoch": 2.844167408726625, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.59837532043457, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8724287748336792, "num_tokens": 853152776.0, "step": 22358 }, { "epoch": 2.8442946190052156, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.4979887008667, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8726959228515625, "num_tokens": 853194176.0, "step": 22359 }, { "epoch": 2.844421829283806, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.596760749816895, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8720493316650391, "num_tokens": 853236834.0, "step": 22360 }, { "epoch": 2.8445490395623967, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.628804206848145, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8668719530105591, "num_tokens": 853274145.0, "step": 22361 }, { "epoch": 2.844676249840987, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61237621307373, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8777966499328613, "num_tokens": 853312569.0, "step": 22362 }, { "epoch": 2.8448034601195777, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.613025665283203, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8793715834617615, "num_tokens": 853351742.0, "step": 22363 }, { "epoch": 2.8449306703981683, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.520197868347168, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8770749568939209, "num_tokens": 853389607.0, "step": 22364 }, { "epoch": 2.845057880676759, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.650009155273438, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8818060755729675, "num_tokens": 853427218.0, "step": 22365 }, { "epoch": 2.8451850909553493, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.572896957397461, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8785107731819153, "num_tokens": 853464773.0, "step": 22366 }, { "epoch": 2.84531230123394, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.599886894226074, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8747596144676208, "num_tokens": 853507250.0, "step": 22367 }, { "epoch": 2.84543951151253, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.573431968688965, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8766336441040039, "num_tokens": 853551947.0, "step": 22368 }, { "epoch": 2.845566721791121, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.61722469329834, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8667935132980347, "num_tokens": 853590557.0, "step": 22369 }, { "epoch": 2.845693932069711, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.607070922851562, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8687635660171509, "num_tokens": 853630870.0, "step": 22370 }, { "epoch": 2.845821142348302, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.520317077636719, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8727415800094604, "num_tokens": 853662360.0, "step": 22371 }, { "epoch": 2.845948352626892, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.658080101013184, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8687598705291748, "num_tokens": 853703661.0, "step": 22372 }, { "epoch": 2.846075562905483, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.4656343460083, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8648998737335205, "num_tokens": 853742398.0, "step": 22373 }, { "epoch": 2.846202773184073, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.626651763916016, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8812393546104431, "num_tokens": 853781479.0, "step": 22374 }, { "epoch": 2.8463299834626636, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.57386302947998, "learning_rate": 1e-06, "loss": 0.4736, "mean_token_accuracy": 0.8497576117515564, "num_tokens": 853822330.0, "step": 22375 }, { "epoch": 2.846457193741254, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.572794914245605, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8626677989959717, "num_tokens": 853862747.0, "step": 22376 }, { "epoch": 2.8465844040198447, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.598274230957031, "learning_rate": 1e-06, "loss": 0.328, "mean_token_accuracy": 0.8969060778617859, "num_tokens": 853897124.0, "step": 22377 }, { "epoch": 2.8467116142984352, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.6217679977417, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8610835075378418, "num_tokens": 853930516.0, "step": 22378 }, { "epoch": 2.8468388245770258, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.582414627075195, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8792730569839478, "num_tokens": 853967016.0, "step": 22379 }, { "epoch": 2.8469660348556163, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.54377269744873, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8578636646270752, "num_tokens": 854008804.0, "step": 22380 }, { "epoch": 2.847093245134207, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.630753517150879, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8616222143173218, "num_tokens": 854048393.0, "step": 22381 }, { "epoch": 2.8472204554127973, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.58503246307373, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8794794082641602, "num_tokens": 854089480.0, "step": 22382 }, { "epoch": 2.847347665691388, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.551408767700195, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8861907720565796, "num_tokens": 854125677.0, "step": 22383 }, { "epoch": 2.8474748759699784, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.649940490722656, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8692890405654907, "num_tokens": 854160485.0, "step": 22384 }, { "epoch": 2.847602086248569, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55406665802002, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8665116429328918, "num_tokens": 854200635.0, "step": 22385 }, { "epoch": 2.8477292965271594, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.646645545959473, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8783710598945618, "num_tokens": 854240046.0, "step": 22386 }, { "epoch": 2.84785650680575, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.548038482666016, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8846201300621033, "num_tokens": 854275880.0, "step": 22387 }, { "epoch": 2.8479837170843405, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.624195098876953, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8761026859283447, "num_tokens": 854311204.0, "step": 22388 }, { "epoch": 2.848110927362931, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.625648498535156, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8714978694915771, "num_tokens": 854346076.0, "step": 22389 }, { "epoch": 2.8482381376415216, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.561851501464844, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8814190626144409, "num_tokens": 854383075.0, "step": 22390 }, { "epoch": 2.848365347920112, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.607404708862305, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8671308755874634, "num_tokens": 854416011.0, "step": 22391 }, { "epoch": 2.8484925581987026, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.681256294250488, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.875998854637146, "num_tokens": 854449520.0, "step": 22392 }, { "epoch": 2.8486197684772927, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.526004791259766, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8729028701782227, "num_tokens": 854484355.0, "step": 22393 }, { "epoch": 2.8487469787558837, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.587912559509277, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8604128956794739, "num_tokens": 854526238.0, "step": 22394 }, { "epoch": 2.8488741890344738, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.49902057647705, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8770349025726318, "num_tokens": 854562619.0, "step": 22395 }, { "epoch": 2.8490013993130647, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.59761905670166, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8792613744735718, "num_tokens": 854604974.0, "step": 22396 }, { "epoch": 2.849128609591655, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.533286094665527, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8806461691856384, "num_tokens": 854644066.0, "step": 22397 }, { "epoch": 2.849255819870246, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.616660118103027, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.876976490020752, "num_tokens": 854682723.0, "step": 22398 }, { "epoch": 2.849383030148836, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.592630386352539, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8849565386772156, "num_tokens": 854720292.0, "step": 22399 }, { "epoch": 2.8495102404274264, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.605185508728027, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8700939416885376, "num_tokens": 854756917.0, "step": 22400 }, { "epoch": 2.849637450706017, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.518843650817871, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8760551810264587, "num_tokens": 854798771.0, "step": 22401 }, { "epoch": 2.8497646609846075, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.623709678649902, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.871777355670929, "num_tokens": 854838193.0, "step": 22402 }, { "epoch": 2.849891871263198, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.570150375366211, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8689879179000854, "num_tokens": 854871650.0, "step": 22403 }, { "epoch": 2.8500190815417885, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.534553527832031, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8736990094184875, "num_tokens": 854910757.0, "step": 22404 }, { "epoch": 2.850146291820379, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.58566951751709, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8625997304916382, "num_tokens": 854945018.0, "step": 22405 }, { "epoch": 2.8502735020989696, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.549019813537598, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8669772744178772, "num_tokens": 854988369.0, "step": 22406 }, { "epoch": 2.85040071237756, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.60496997833252, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8765756487846375, "num_tokens": 855023935.0, "step": 22407 }, { "epoch": 2.8505279226561506, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.578204154968262, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8738577365875244, "num_tokens": 855062002.0, "step": 22408 }, { "epoch": 2.850655132934741, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.547586441040039, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8773069381713867, "num_tokens": 855101257.0, "step": 22409 }, { "epoch": 2.8507823432133317, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.560454368591309, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8756304979324341, "num_tokens": 855135834.0, "step": 22410 }, { "epoch": 2.850909553491922, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.694401741027832, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8767493963241577, "num_tokens": 855174156.0, "step": 22411 }, { "epoch": 2.8510367637705127, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.551057815551758, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.884217381477356, "num_tokens": 855207698.0, "step": 22412 }, { "epoch": 2.8511639740491033, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.63907527923584, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8753020763397217, "num_tokens": 855240636.0, "step": 22413 }, { "epoch": 2.851291184327694, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.52753734588623, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8773331642150879, "num_tokens": 855279923.0, "step": 22414 }, { "epoch": 2.8514183946062843, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.651734352111816, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8629987239837646, "num_tokens": 855315460.0, "step": 22415 }, { "epoch": 2.851545604884875, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.536948204040527, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8819527626037598, "num_tokens": 855351548.0, "step": 22416 }, { "epoch": 2.8516728151634654, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.654726028442383, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8569972515106201, "num_tokens": 855390715.0, "step": 22417 }, { "epoch": 2.8518000254420555, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.537002563476562, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8615759611129761, "num_tokens": 855431997.0, "step": 22418 }, { "epoch": 2.8519272357206464, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.603782653808594, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8685036897659302, "num_tokens": 855470242.0, "step": 22419 }, { "epoch": 2.8520544459992365, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.624541282653809, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8618510961532593, "num_tokens": 855508901.0, "step": 22420 }, { "epoch": 2.8521816562778275, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.629220008850098, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8702530264854431, "num_tokens": 855545290.0, "step": 22421 }, { "epoch": 2.8523088665564176, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.570430755615234, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.872403621673584, "num_tokens": 855579567.0, "step": 22422 }, { "epoch": 2.852436076835008, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.601554870605469, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8671011328697205, "num_tokens": 855621117.0, "step": 22423 }, { "epoch": 2.8525632871135986, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.56081771850586, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8762550354003906, "num_tokens": 855660494.0, "step": 22424 }, { "epoch": 2.852690497392189, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.608282089233398, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8737677335739136, "num_tokens": 855694043.0, "step": 22425 }, { "epoch": 2.8528177076707797, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.55411434173584, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8601062297821045, "num_tokens": 855734615.0, "step": 22426 }, { "epoch": 2.85294491794937, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.572957038879395, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.884667694568634, "num_tokens": 855771141.0, "step": 22427 }, { "epoch": 2.8530721282279607, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.66659927368164, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8739842772483826, "num_tokens": 855806877.0, "step": 22428 }, { "epoch": 2.8531993385065513, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.622162818908691, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8884443044662476, "num_tokens": 855840076.0, "step": 22429 }, { "epoch": 2.853326548785142, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.604324340820312, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8844510912895203, "num_tokens": 855877972.0, "step": 22430 }, { "epoch": 2.8534537590637323, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.660555839538574, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8691685199737549, "num_tokens": 855917424.0, "step": 22431 }, { "epoch": 2.853580969342323, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.544027328491211, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8636348843574524, "num_tokens": 855954853.0, "step": 22432 }, { "epoch": 2.8537081796209134, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.624028205871582, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8757349848747253, "num_tokens": 855993300.0, "step": 22433 }, { "epoch": 2.853835389899504, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.596396446228027, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8778349161148071, "num_tokens": 856035307.0, "step": 22434 }, { "epoch": 2.8539626001780944, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.509739875793457, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8719471096992493, "num_tokens": 856071801.0, "step": 22435 }, { "epoch": 2.854089810456685, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.625237464904785, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.862075686454773, "num_tokens": 856112035.0, "step": 22436 }, { "epoch": 2.8542170207352755, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.528182983398438, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8802151083946228, "num_tokens": 856148275.0, "step": 22437 }, { "epoch": 2.854344231013866, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.647873878479004, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8774865865707397, "num_tokens": 856180330.0, "step": 22438 }, { "epoch": 2.8544714412924566, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.580363273620605, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8614409565925598, "num_tokens": 856219354.0, "step": 22439 }, { "epoch": 2.854598651571047, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.578653335571289, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8842991590499878, "num_tokens": 856257583.0, "step": 22440 }, { "epoch": 2.8547258618496376, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.651928901672363, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8568623661994934, "num_tokens": 856298056.0, "step": 22441 }, { "epoch": 2.854853072128228, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.584305763244629, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8712429404258728, "num_tokens": 856334109.0, "step": 22442 }, { "epoch": 2.8549802824068182, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.604480743408203, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.876206636428833, "num_tokens": 856376655.0, "step": 22443 }, { "epoch": 2.855107492685409, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.578320503234863, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.8898779153823853, "num_tokens": 856409926.0, "step": 22444 }, { "epoch": 2.8552347029639993, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.60954475402832, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8659707903862, "num_tokens": 856451471.0, "step": 22445 }, { "epoch": 2.8553619132425903, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.658453941345215, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8825550675392151, "num_tokens": 856483237.0, "step": 22446 }, { "epoch": 2.8554891235211803, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.595869064331055, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8655916452407837, "num_tokens": 856521278.0, "step": 22447 }, { "epoch": 2.855616333799771, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.631426811218262, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.871364414691925, "num_tokens": 856565048.0, "step": 22448 }, { "epoch": 2.8557435440783614, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.583640098571777, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.87752765417099, "num_tokens": 856603756.0, "step": 22449 }, { "epoch": 2.855870754356952, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.625990867614746, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8799072504043579, "num_tokens": 856640677.0, "step": 22450 }, { "epoch": 2.8559979646355425, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.518787384033203, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8743669390678406, "num_tokens": 856679982.0, "step": 22451 }, { "epoch": 2.856125174914133, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.540699005126953, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8633173704147339, "num_tokens": 856720621.0, "step": 22452 }, { "epoch": 2.8562523851927235, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.540935516357422, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8822424411773682, "num_tokens": 856757172.0, "step": 22453 }, { "epoch": 2.856379595471314, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.514741897583008, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8711340427398682, "num_tokens": 856796876.0, "step": 22454 }, { "epoch": 2.8565068057499046, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.6915864944458, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8739974498748779, "num_tokens": 856831148.0, "step": 22455 }, { "epoch": 2.856634016028495, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.512053489685059, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8704614639282227, "num_tokens": 856867091.0, "step": 22456 }, { "epoch": 2.8567612263070856, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.53416919708252, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.864657461643219, "num_tokens": 856906118.0, "step": 22457 }, { "epoch": 2.856888436585676, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.615038871765137, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8490667343139648, "num_tokens": 856942694.0, "step": 22458 }, { "epoch": 2.8570156468642667, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.61271858215332, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8661976456642151, "num_tokens": 856977530.0, "step": 22459 }, { "epoch": 2.857142857142857, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.560905456542969, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8773972988128662, "num_tokens": 857012417.0, "step": 22460 }, { "epoch": 2.8572700674214477, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.59068489074707, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8670204877853394, "num_tokens": 857053556.0, "step": 22461 }, { "epoch": 2.8573972777000383, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.623285293579102, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8772681951522827, "num_tokens": 857094514.0, "step": 22462 }, { "epoch": 2.857524487978629, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.552714347839355, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8754361271858215, "num_tokens": 857133670.0, "step": 22463 }, { "epoch": 2.8576516982572193, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.61658000946045, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8671795129776001, "num_tokens": 857167934.0, "step": 22464 }, { "epoch": 2.85777890853581, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.561092376708984, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8575764894485474, "num_tokens": 857208263.0, "step": 22465 }, { "epoch": 2.8579061188144, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.491273880004883, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.856684684753418, "num_tokens": 857255107.0, "step": 22466 }, { "epoch": 2.858033329092991, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.543815612792969, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8710026144981384, "num_tokens": 857298368.0, "step": 22467 }, { "epoch": 2.858160539371581, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.562774658203125, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.874588668346405, "num_tokens": 857336899.0, "step": 22468 }, { "epoch": 2.858287749650172, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.526025772094727, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8783019781112671, "num_tokens": 857378482.0, "step": 22469 }, { "epoch": 2.858414959928762, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.554253578186035, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8831304907798767, "num_tokens": 857413203.0, "step": 22470 }, { "epoch": 2.858542170207353, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.547438621520996, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8726120591163635, "num_tokens": 857446706.0, "step": 22471 }, { "epoch": 2.858669380485943, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.536108016967773, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8773082494735718, "num_tokens": 857483415.0, "step": 22472 }, { "epoch": 2.8587965907645336, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.495436668395996, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.879224419593811, "num_tokens": 857525323.0, "step": 22473 }, { "epoch": 2.858923801043124, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.548633575439453, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8901591897010803, "num_tokens": 857564400.0, "step": 22474 }, { "epoch": 2.8590510113217147, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.545738220214844, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8728808760643005, "num_tokens": 857602756.0, "step": 22475 }, { "epoch": 2.859178221600305, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.602693557739258, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8739845156669617, "num_tokens": 857636875.0, "step": 22476 }, { "epoch": 2.8593054318788957, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.521734237670898, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8910022974014282, "num_tokens": 857671950.0, "step": 22477 }, { "epoch": 2.8594326421574863, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.559513092041016, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8864260911941528, "num_tokens": 857705274.0, "step": 22478 }, { "epoch": 2.859559852436077, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.541616439819336, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8680083751678467, "num_tokens": 857743730.0, "step": 22479 }, { "epoch": 2.8596870627146673, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.545519828796387, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8726375102996826, "num_tokens": 857784055.0, "step": 22480 }, { "epoch": 2.859814272993258, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.53707218170166, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8808848261833191, "num_tokens": 857822431.0, "step": 22481 }, { "epoch": 2.8599414832718484, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.528746604919434, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8861964344978333, "num_tokens": 857859258.0, "step": 22482 }, { "epoch": 2.860068693550439, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.648540496826172, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8657898902893066, "num_tokens": 857901209.0, "step": 22483 }, { "epoch": 2.8601959038290294, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.505066871643066, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8870615363121033, "num_tokens": 857935563.0, "step": 22484 }, { "epoch": 2.86032311410762, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.5943603515625, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8654319047927856, "num_tokens": 857966777.0, "step": 22485 }, { "epoch": 2.8604503243862105, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.555724143981934, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.879132866859436, "num_tokens": 858005078.0, "step": 22486 }, { "epoch": 2.860577534664801, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.516707420349121, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8616687059402466, "num_tokens": 858049288.0, "step": 22487 }, { "epoch": 2.8607047449433916, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.5870943069458, "learning_rate": 1e-06, "loss": 0.3406, "mean_token_accuracy": 0.8938050270080566, "num_tokens": 858089834.0, "step": 22488 }, { "epoch": 2.860831955221982, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.521307945251465, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8791958093643188, "num_tokens": 858124152.0, "step": 22489 }, { "epoch": 2.8609591655005726, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.570711135864258, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8794506788253784, "num_tokens": 858163202.0, "step": 22490 }, { "epoch": 2.8610863757791627, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.546700477600098, "learning_rate": 1e-06, "loss": 0.3371, "mean_token_accuracy": 0.8934227228164673, "num_tokens": 858199173.0, "step": 22491 }, { "epoch": 2.8612135860577537, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.561697006225586, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8723853826522827, "num_tokens": 858239693.0, "step": 22492 }, { "epoch": 2.8613407963363438, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.564027786254883, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8818468451499939, "num_tokens": 858280720.0, "step": 22493 }, { "epoch": 2.8614680066149347, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.538875579833984, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8864943385124207, "num_tokens": 858314145.0, "step": 22494 }, { "epoch": 2.861595216893525, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.574495315551758, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8764524459838867, "num_tokens": 858355177.0, "step": 22495 }, { "epoch": 2.861722427172116, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.542474746704102, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8785691261291504, "num_tokens": 858393070.0, "step": 22496 }, { "epoch": 2.861849637450706, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.561094284057617, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8749507665634155, "num_tokens": 858431827.0, "step": 22497 }, { "epoch": 2.8619768477292964, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.564251899719238, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8715450763702393, "num_tokens": 858474990.0, "step": 22498 }, { "epoch": 2.862104058007887, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.56811237335205, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8875945806503296, "num_tokens": 858514698.0, "step": 22499 }, { "epoch": 2.8622312682864774, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.537737846374512, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8758021593093872, "num_tokens": 858557601.0, "step": 22500 }, { "epoch": 2.862358478565068, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.590624809265137, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8829227685928345, "num_tokens": 858596830.0, "step": 22501 }, { "epoch": 2.8624856888436585, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.585603713989258, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8653783202171326, "num_tokens": 858637555.0, "step": 22502 }, { "epoch": 2.862612899122249, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.52904987335205, "learning_rate": 1e-06, "loss": 0.3357, "mean_token_accuracy": 0.8936804533004761, "num_tokens": 858672813.0, "step": 22503 }, { "epoch": 2.8627401094008396, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.582680702209473, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8808534145355225, "num_tokens": 858710743.0, "step": 22504 }, { "epoch": 2.86286731967943, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.594356536865234, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8867461681365967, "num_tokens": 858739486.0, "step": 22505 }, { "epoch": 2.8629945299580206, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.552210807800293, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8549081087112427, "num_tokens": 858780678.0, "step": 22506 }, { "epoch": 2.863121740236611, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.511994361877441, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.870294451713562, "num_tokens": 858826112.0, "step": 22507 }, { "epoch": 2.8632489505152017, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.597593307495117, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8857520818710327, "num_tokens": 858862158.0, "step": 22508 }, { "epoch": 2.863376160793792, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.611189842224121, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8790315389633179, "num_tokens": 858905270.0, "step": 22509 }, { "epoch": 2.8635033710723827, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.603086471557617, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8681017160415649, "num_tokens": 858941802.0, "step": 22510 }, { "epoch": 2.8636305813509733, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.556370735168457, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8778319358825684, "num_tokens": 858974672.0, "step": 22511 }, { "epoch": 2.863757791629564, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.557478904724121, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8730267286300659, "num_tokens": 859015305.0, "step": 22512 }, { "epoch": 2.8638850019081543, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.553494453430176, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8658320307731628, "num_tokens": 859050646.0, "step": 22513 }, { "epoch": 2.864012212186745, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.601884841918945, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8845019340515137, "num_tokens": 859090139.0, "step": 22514 }, { "epoch": 2.8641394224653354, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.562614440917969, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8788406252861023, "num_tokens": 859130661.0, "step": 22515 }, { "epoch": 2.8642666327439255, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.628173828125, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8748972415924072, "num_tokens": 859166399.0, "step": 22516 }, { "epoch": 2.8643938430225164, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.52392578125, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8751318454742432, "num_tokens": 859208349.0, "step": 22517 }, { "epoch": 2.8645210533011065, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.591739654541016, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.878423273563385, "num_tokens": 859243784.0, "step": 22518 }, { "epoch": 2.8646482635796975, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.57409381866455, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.86101233959198, "num_tokens": 859278345.0, "step": 22519 }, { "epoch": 2.8647754738582876, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.539754867553711, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8652783036231995, "num_tokens": 859317454.0, "step": 22520 }, { "epoch": 2.864902684136878, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.578086853027344, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8702389001846313, "num_tokens": 859352076.0, "step": 22521 }, { "epoch": 2.8650298944154686, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.550841331481934, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8832798600196838, "num_tokens": 859389957.0, "step": 22522 }, { "epoch": 2.865157104694059, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.58138370513916, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8748466968536377, "num_tokens": 859425167.0, "step": 22523 }, { "epoch": 2.8652843149726497, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.55880355834961, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.871425986289978, "num_tokens": 859470999.0, "step": 22524 }, { "epoch": 2.86541152525124, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.607722282409668, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8818174600601196, "num_tokens": 859503763.0, "step": 22525 }, { "epoch": 2.8655387355298307, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.577140808105469, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.883437991142273, "num_tokens": 859537739.0, "step": 22526 }, { "epoch": 2.8656659458084213, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.591012954711914, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8834619522094727, "num_tokens": 859573497.0, "step": 22527 }, { "epoch": 2.865793156087012, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.569145202636719, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8710650205612183, "num_tokens": 859606997.0, "step": 22528 }, { "epoch": 2.8659203663656023, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.628804206848145, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8634254932403564, "num_tokens": 859642278.0, "step": 22529 }, { "epoch": 2.866047576644193, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.524407386779785, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8868096470832825, "num_tokens": 859684728.0, "step": 22530 }, { "epoch": 2.8661747869227834, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.626057624816895, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8684097528457642, "num_tokens": 859720946.0, "step": 22531 }, { "epoch": 2.866301997201374, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.586880683898926, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8730286359786987, "num_tokens": 859752605.0, "step": 22532 }, { "epoch": 2.8664292074799644, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.562629699707031, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8878097534179688, "num_tokens": 859784206.0, "step": 22533 }, { "epoch": 2.866556417758555, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.529507637023926, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8710708618164062, "num_tokens": 859828142.0, "step": 22534 }, { "epoch": 2.8666836280371455, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.562044143676758, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8700516223907471, "num_tokens": 859870032.0, "step": 22535 }, { "epoch": 2.866810838315736, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.576848030090332, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8652498722076416, "num_tokens": 859911417.0, "step": 22536 }, { "epoch": 2.8669380485943265, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.662489891052246, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8854358196258545, "num_tokens": 859953929.0, "step": 22537 }, { "epoch": 2.867065258872917, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.487884521484375, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.879128098487854, "num_tokens": 859985402.0, "step": 22538 }, { "epoch": 2.8671924691515076, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.6503324508667, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8725007176399231, "num_tokens": 860017374.0, "step": 22539 }, { "epoch": 2.867319679430098, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.547161102294922, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8843871355056763, "num_tokens": 860058217.0, "step": 22540 }, { "epoch": 2.867446889708688, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.652387619018555, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8585968613624573, "num_tokens": 860092317.0, "step": 22541 }, { "epoch": 2.867574099987279, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.634542465209961, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8716926574707031, "num_tokens": 860125067.0, "step": 22542 }, { "epoch": 2.8677013102658693, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.584501266479492, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8871139287948608, "num_tokens": 860163805.0, "step": 22543 }, { "epoch": 2.8678285205444602, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.625476837158203, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8629186749458313, "num_tokens": 860200178.0, "step": 22544 }, { "epoch": 2.8679557308230503, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.578794479370117, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8743690848350525, "num_tokens": 860235283.0, "step": 22545 }, { "epoch": 2.868082941101641, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.577458381652832, "learning_rate": 1e-06, "loss": 0.3444, "mean_token_accuracy": 0.8911113739013672, "num_tokens": 860276776.0, "step": 22546 }, { "epoch": 2.8682101513802314, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.517297744750977, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8792688846588135, "num_tokens": 860318492.0, "step": 22547 }, { "epoch": 2.868337361658822, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.712010383605957, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8602182865142822, "num_tokens": 860358556.0, "step": 22548 }, { "epoch": 2.8684645719374124, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.580754280090332, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8658081293106079, "num_tokens": 860391114.0, "step": 22549 }, { "epoch": 2.868591782216003, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.628568649291992, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8709911704063416, "num_tokens": 860430696.0, "step": 22550 }, { "epoch": 2.8687189924945935, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.605979919433594, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8821009397506714, "num_tokens": 860471055.0, "step": 22551 }, { "epoch": 2.868846202773184, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.625109672546387, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8677276372909546, "num_tokens": 860512472.0, "step": 22552 }, { "epoch": 2.8689734130517746, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.641670227050781, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8696485161781311, "num_tokens": 860541014.0, "step": 22553 }, { "epoch": 2.869100623330365, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.572270393371582, "learning_rate": 1e-06, "loss": 0.3537, "mean_token_accuracy": 0.888961911201477, "num_tokens": 860585025.0, "step": 22554 }, { "epoch": 2.8692278336089556, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.62350845336914, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8832789063453674, "num_tokens": 860616361.0, "step": 22555 }, { "epoch": 2.869355043887546, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.665639877319336, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8669114112854004, "num_tokens": 860655976.0, "step": 22556 }, { "epoch": 2.8694822541661367, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.564440727233887, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8633494973182678, "num_tokens": 860703949.0, "step": 22557 }, { "epoch": 2.869609464444727, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.6210355758667, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8745995759963989, "num_tokens": 860739935.0, "step": 22558 }, { "epoch": 2.8697366747233177, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.616658210754395, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8783953189849854, "num_tokens": 860782483.0, "step": 22559 }, { "epoch": 2.8698638850019083, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.63769245147705, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8640841245651245, "num_tokens": 860816821.0, "step": 22560 }, { "epoch": 2.869991095280499, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.676058769226074, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8636672496795654, "num_tokens": 860851313.0, "step": 22561 }, { "epoch": 2.8701183055590893, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.578819274902344, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8831813335418701, "num_tokens": 860890540.0, "step": 22562 }, { "epoch": 2.87024551583768, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.638752937316895, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8689112663269043, "num_tokens": 860929686.0, "step": 22563 }, { "epoch": 2.87037272611627, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.636834144592285, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8777661919593811, "num_tokens": 860968186.0, "step": 22564 }, { "epoch": 2.870499936394861, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.601513862609863, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.871485710144043, "num_tokens": 861007252.0, "step": 22565 }, { "epoch": 2.870627146673451, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.545600891113281, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8761388063430786, "num_tokens": 861049898.0, "step": 22566 }, { "epoch": 2.870754356952042, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.664551734924316, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8821302652359009, "num_tokens": 861087711.0, "step": 22567 }, { "epoch": 2.870881567230632, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.607142448425293, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8550000190734863, "num_tokens": 861128566.0, "step": 22568 }, { "epoch": 2.871008777509223, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.679798126220703, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8788440227508545, "num_tokens": 861156299.0, "step": 22569 }, { "epoch": 2.871135987787813, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.650421142578125, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8858301639556885, "num_tokens": 861189376.0, "step": 22570 }, { "epoch": 2.8712631980664036, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.648714065551758, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8717043995857239, "num_tokens": 861227156.0, "step": 22571 }, { "epoch": 2.871390408344994, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.67366886138916, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8742877244949341, "num_tokens": 861268088.0, "step": 22572 }, { "epoch": 2.8715176186235847, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.635722160339355, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.877947986125946, "num_tokens": 861308892.0, "step": 22573 }, { "epoch": 2.871644828902175, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.59666919708252, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8770651817321777, "num_tokens": 861350108.0, "step": 22574 }, { "epoch": 2.8717720391807657, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.631794929504395, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8841473460197449, "num_tokens": 861390553.0, "step": 22575 }, { "epoch": 2.8718992494593563, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.650123596191406, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8774150013923645, "num_tokens": 861425878.0, "step": 22576 }, { "epoch": 2.872026459737947, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.528342247009277, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8910795450210571, "num_tokens": 861460935.0, "step": 22577 }, { "epoch": 2.8721536700165373, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.630833625793457, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8585513830184937, "num_tokens": 861500446.0, "step": 22578 }, { "epoch": 2.872280880295128, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.551084518432617, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8631001114845276, "num_tokens": 861538000.0, "step": 22579 }, { "epoch": 2.8724080905737184, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.696521759033203, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8775413036346436, "num_tokens": 861578052.0, "step": 22580 }, { "epoch": 2.872535300852309, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.624435424804688, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8707181811332703, "num_tokens": 861619065.0, "step": 22581 }, { "epoch": 2.8726625111308994, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.634775161743164, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8783135414123535, "num_tokens": 861660112.0, "step": 22582 }, { "epoch": 2.87278972140949, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.647584915161133, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8763613104820251, "num_tokens": 861699312.0, "step": 22583 }, { "epoch": 2.8729169316880805, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.558743476867676, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8917808532714844, "num_tokens": 861735167.0, "step": 22584 }, { "epoch": 2.873044141966671, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.640922546386719, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8689863681793213, "num_tokens": 861778594.0, "step": 22585 }, { "epoch": 2.8731713522452615, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.58670711517334, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8740624785423279, "num_tokens": 861814754.0, "step": 22586 }, { "epoch": 2.873298562523852, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.597186088562012, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8607773780822754, "num_tokens": 861849019.0, "step": 22587 }, { "epoch": 2.8734257728024426, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.512872695922852, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8712654113769531, "num_tokens": 861889471.0, "step": 22588 }, { "epoch": 2.8735529830810327, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.611139297485352, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.866668164730072, "num_tokens": 861926811.0, "step": 22589 }, { "epoch": 2.8736801933596237, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.575053215026855, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.872620701789856, "num_tokens": 861963628.0, "step": 22590 }, { "epoch": 2.8738074036382137, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.654460906982422, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8632394075393677, "num_tokens": 862002915.0, "step": 22591 }, { "epoch": 2.8739346139168047, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.577953338623047, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8810803890228271, "num_tokens": 862047276.0, "step": 22592 }, { "epoch": 2.874061824195395, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.612569808959961, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8825902938842773, "num_tokens": 862091241.0, "step": 22593 }, { "epoch": 2.8741890344739858, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.585159301757812, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8776083588600159, "num_tokens": 862126025.0, "step": 22594 }, { "epoch": 2.874316244752576, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.67344856262207, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8655176162719727, "num_tokens": 862160822.0, "step": 22595 }, { "epoch": 2.8744434550311664, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.589800834655762, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8793812394142151, "num_tokens": 862196808.0, "step": 22596 }, { "epoch": 2.874570665309757, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.577654838562012, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.860604465007782, "num_tokens": 862233722.0, "step": 22597 }, { "epoch": 2.8746978755883474, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.700787544250488, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8899332284927368, "num_tokens": 862271914.0, "step": 22598 }, { "epoch": 2.874825085866938, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.562902450561523, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8711746335029602, "num_tokens": 862313696.0, "step": 22599 }, { "epoch": 2.8749522961455285, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.601984977722168, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8683257699012756, "num_tokens": 862355366.0, "step": 22600 }, { "epoch": 2.875079506424119, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.526708602905273, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8653442859649658, "num_tokens": 862395630.0, "step": 22601 }, { "epoch": 2.8752067167027096, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.615591049194336, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8757717609405518, "num_tokens": 862429650.0, "step": 22602 }, { "epoch": 2.8753339269813, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.575030326843262, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8714624047279358, "num_tokens": 862470921.0, "step": 22603 }, { "epoch": 2.8754611372598906, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.565290451049805, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8804682493209839, "num_tokens": 862505204.0, "step": 22604 }, { "epoch": 2.875588347538481, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.557084083557129, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8737199306488037, "num_tokens": 862548686.0, "step": 22605 }, { "epoch": 2.8757155578170717, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.623600006103516, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8873319625854492, "num_tokens": 862585920.0, "step": 22606 }, { "epoch": 2.875842768095662, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.610563278198242, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.87013840675354, "num_tokens": 862619364.0, "step": 22607 }, { "epoch": 2.8759699783742527, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.651215553283691, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8709260821342468, "num_tokens": 862661317.0, "step": 22608 }, { "epoch": 2.8760971886528433, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.607804298400879, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.877983033657074, "num_tokens": 862698236.0, "step": 22609 }, { "epoch": 2.876224398931434, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.578100204467773, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8850708603858948, "num_tokens": 862739224.0, "step": 22610 }, { "epoch": 2.8763516092100243, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.595122337341309, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.889250636100769, "num_tokens": 862776534.0, "step": 22611 }, { "epoch": 2.876478819488615, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.572283744812012, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8681469559669495, "num_tokens": 862815690.0, "step": 22612 }, { "epoch": 2.8766060297672054, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.6698637008667, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8737332820892334, "num_tokens": 862851021.0, "step": 22613 }, { "epoch": 2.8767332400457954, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.585698127746582, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8827778100967407, "num_tokens": 862885122.0, "step": 22614 }, { "epoch": 2.8768604503243864, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.66085433959961, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8813678622245789, "num_tokens": 862922233.0, "step": 22615 }, { "epoch": 2.8769876606029765, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.686444282531738, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8845894932746887, "num_tokens": 862961355.0, "step": 22616 }, { "epoch": 2.8771148708815675, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.504731178283691, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.882868766784668, "num_tokens": 862998523.0, "step": 22617 }, { "epoch": 2.8772420811601576, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.665173530578613, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8756918907165527, "num_tokens": 863039684.0, "step": 22618 }, { "epoch": 2.877369291438748, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.5917387008667, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8792572021484375, "num_tokens": 863077649.0, "step": 22619 }, { "epoch": 2.8774965017173386, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.571866989135742, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8708752393722534, "num_tokens": 863116816.0, "step": 22620 }, { "epoch": 2.877623711995929, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.640886306762695, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8857093453407288, "num_tokens": 863155306.0, "step": 22621 }, { "epoch": 2.8777509222745197, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.555110931396484, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8826820254325867, "num_tokens": 863194381.0, "step": 22622 }, { "epoch": 2.87787813255311, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.628631591796875, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8740302920341492, "num_tokens": 863231470.0, "step": 22623 }, { "epoch": 2.8780053428317007, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.563114166259766, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8676290512084961, "num_tokens": 863265617.0, "step": 22624 }, { "epoch": 2.8781325531102913, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.594732284545898, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.864844799041748, "num_tokens": 863300111.0, "step": 22625 }, { "epoch": 2.878259763388882, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.589125633239746, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.882522463798523, "num_tokens": 863338626.0, "step": 22626 }, { "epoch": 2.8783869736674723, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.562585830688477, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8671761751174927, "num_tokens": 863374672.0, "step": 22627 }, { "epoch": 2.878514183946063, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.516646385192871, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.870477557182312, "num_tokens": 863412717.0, "step": 22628 }, { "epoch": 2.8786413942246534, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.552491188049316, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.88303142786026, "num_tokens": 863447776.0, "step": 22629 }, { "epoch": 2.878768604503244, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.627521514892578, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.8916759490966797, "num_tokens": 863485697.0, "step": 22630 }, { "epoch": 2.8788958147818344, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.541396141052246, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.870307207107544, "num_tokens": 863528706.0, "step": 22631 }, { "epoch": 2.879023025060425, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.63387680053711, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8566640019416809, "num_tokens": 863569181.0, "step": 22632 }, { "epoch": 2.8791502353390155, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.597028732299805, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8600641489028931, "num_tokens": 863602985.0, "step": 22633 }, { "epoch": 2.879277445617606, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.635626792907715, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8794187307357788, "num_tokens": 863634590.0, "step": 22634 }, { "epoch": 2.8794046558961965, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.638924598693848, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8733943700790405, "num_tokens": 863677671.0, "step": 22635 }, { "epoch": 2.879531866174787, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.521138191223145, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8770430088043213, "num_tokens": 863716102.0, "step": 22636 }, { "epoch": 2.8796590764533776, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.700271606445312, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8674310445785522, "num_tokens": 863751560.0, "step": 22637 }, { "epoch": 2.879786286731968, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.592268943786621, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8893365859985352, "num_tokens": 863792373.0, "step": 22638 }, { "epoch": 2.879913497010558, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.594666481018066, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8704545497894287, "num_tokens": 863829969.0, "step": 22639 }, { "epoch": 2.880040707289149, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.555770874023438, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8802359104156494, "num_tokens": 863867080.0, "step": 22640 }, { "epoch": 2.8801679175677393, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.59821891784668, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8791317939758301, "num_tokens": 863904230.0, "step": 22641 }, { "epoch": 2.8802951278463302, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.557771682739258, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.879575788974762, "num_tokens": 863947309.0, "step": 22642 }, { "epoch": 2.8804223381249203, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.598160743713379, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8803389668464661, "num_tokens": 863985557.0, "step": 22643 }, { "epoch": 2.880549548403511, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.582121849060059, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8634819388389587, "num_tokens": 864021842.0, "step": 22644 }, { "epoch": 2.8806767586821014, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.543699264526367, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8670650124549866, "num_tokens": 864060765.0, "step": 22645 }, { "epoch": 2.880803968960692, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.618220329284668, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8774148225784302, "num_tokens": 864100624.0, "step": 22646 }, { "epoch": 2.8809311792392824, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.637554168701172, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8851224184036255, "num_tokens": 864136599.0, "step": 22647 }, { "epoch": 2.881058389517873, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.521638870239258, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8779200315475464, "num_tokens": 864169876.0, "step": 22648 }, { "epoch": 2.8811855997964635, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.635683059692383, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8741897344589233, "num_tokens": 864213709.0, "step": 22649 }, { "epoch": 2.881312810075054, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.59410572052002, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8803422451019287, "num_tokens": 864245581.0, "step": 22650 }, { "epoch": 2.8814400203536445, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.67134952545166, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.86066734790802, "num_tokens": 864287081.0, "step": 22651 }, { "epoch": 2.881567230632235, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.600342750549316, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8594111204147339, "num_tokens": 864321480.0, "step": 22652 }, { "epoch": 2.8816944409108256, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.578396797180176, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8814910650253296, "num_tokens": 864358580.0, "step": 22653 }, { "epoch": 2.881821651189416, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.57762336730957, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8659480810165405, "num_tokens": 864396277.0, "step": 22654 }, { "epoch": 2.8819488614680067, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.634492874145508, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8696150779724121, "num_tokens": 864433282.0, "step": 22655 }, { "epoch": 2.882076071746597, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.619121551513672, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8666452169418335, "num_tokens": 864469305.0, "step": 22656 }, { "epoch": 2.8822032820251877, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.57144546508789, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8760025501251221, "num_tokens": 864507246.0, "step": 22657 }, { "epoch": 2.8823304923037782, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.615254402160645, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8688818216323853, "num_tokens": 864547315.0, "step": 22658 }, { "epoch": 2.8824577025823688, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.588583946228027, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8687233328819275, "num_tokens": 864590813.0, "step": 22659 }, { "epoch": 2.8825849128609593, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.573119163513184, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.865795373916626, "num_tokens": 864625078.0, "step": 22660 }, { "epoch": 2.88271212313955, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.658726692199707, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8828608393669128, "num_tokens": 864663714.0, "step": 22661 }, { "epoch": 2.88283933341814, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.576375007629395, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8743757009506226, "num_tokens": 864697026.0, "step": 22662 }, { "epoch": 2.882966543696731, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.625007629394531, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8839246034622192, "num_tokens": 864732337.0, "step": 22663 }, { "epoch": 2.883093753975321, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.563435554504395, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8683067560195923, "num_tokens": 864776598.0, "step": 22664 }, { "epoch": 2.883220964253912, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.629209518432617, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8616673946380615, "num_tokens": 864817279.0, "step": 22665 }, { "epoch": 2.883348174532502, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.633106231689453, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.881729245185852, "num_tokens": 864854914.0, "step": 22666 }, { "epoch": 2.883475384811093, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.656123161315918, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8773417472839355, "num_tokens": 864892914.0, "step": 22667 }, { "epoch": 2.883602595089683, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.61825180053711, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8713979721069336, "num_tokens": 864934189.0, "step": 22668 }, { "epoch": 2.8837298053682736, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.628820419311523, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.88404780626297, "num_tokens": 864972085.0, "step": 22669 }, { "epoch": 2.883857015646864, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.692100524902344, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8583847880363464, "num_tokens": 865008674.0, "step": 22670 }, { "epoch": 2.8839842259254547, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.608667373657227, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8654196858406067, "num_tokens": 865046004.0, "step": 22671 }, { "epoch": 2.884111436204045, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.601180076599121, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8791749477386475, "num_tokens": 865082409.0, "step": 22672 }, { "epoch": 2.8842386464826357, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.662365913391113, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8737965822219849, "num_tokens": 865124457.0, "step": 22673 }, { "epoch": 2.8843658567612263, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.603731155395508, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8861691355705261, "num_tokens": 865162290.0, "step": 22674 }, { "epoch": 2.884493067039817, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.64594554901123, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8731645345687866, "num_tokens": 865201686.0, "step": 22675 }, { "epoch": 2.8846202773184073, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.640837669372559, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8653494119644165, "num_tokens": 865236876.0, "step": 22676 }, { "epoch": 2.884747487596998, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.629484176635742, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8777866363525391, "num_tokens": 865276994.0, "step": 22677 }, { "epoch": 2.8848746978755884, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.678241729736328, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8740773797035217, "num_tokens": 865311990.0, "step": 22678 }, { "epoch": 2.885001908154179, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.676979064941406, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8809171319007874, "num_tokens": 865352324.0, "step": 22679 }, { "epoch": 2.8851291184327694, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.624751091003418, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.874392569065094, "num_tokens": 865391428.0, "step": 22680 }, { "epoch": 2.88525632871136, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.753849983215332, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8730543851852417, "num_tokens": 865428478.0, "step": 22681 }, { "epoch": 2.8853835389899505, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.647383689880371, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.875210702419281, "num_tokens": 865466531.0, "step": 22682 }, { "epoch": 2.885510749268541, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.622150421142578, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8807989954948425, "num_tokens": 865503177.0, "step": 22683 }, { "epoch": 2.8856379595471315, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.69514274597168, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8679009079933167, "num_tokens": 865539454.0, "step": 22684 }, { "epoch": 2.885765169825722, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.653002738952637, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.8926689624786377, "num_tokens": 865571937.0, "step": 22685 }, { "epoch": 2.8858923801043126, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.662222862243652, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8767937421798706, "num_tokens": 865609457.0, "step": 22686 }, { "epoch": 2.8860195903829027, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.646478652954102, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8627681732177734, "num_tokens": 865647391.0, "step": 22687 }, { "epoch": 2.8861468006614936, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.72095775604248, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8735273480415344, "num_tokens": 865680578.0, "step": 22688 }, { "epoch": 2.8862740109400837, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.718453407287598, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8710848093032837, "num_tokens": 865714610.0, "step": 22689 }, { "epoch": 2.8864012212186747, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.655129432678223, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8703715205192566, "num_tokens": 865755672.0, "step": 22690 }, { "epoch": 2.886528431497265, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.717005729675293, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8830525279045105, "num_tokens": 865790111.0, "step": 22691 }, { "epoch": 2.8866556417758558, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.693812370300293, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8753447532653809, "num_tokens": 865825580.0, "step": 22692 }, { "epoch": 2.886782852054446, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.697348594665527, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8779888153076172, "num_tokens": 865869099.0, "step": 22693 }, { "epoch": 2.8869100623330364, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.698404312133789, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8650590181350708, "num_tokens": 865906957.0, "step": 22694 }, { "epoch": 2.887037272611627, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.631967544555664, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8839594125747681, "num_tokens": 865945712.0, "step": 22695 }, { "epoch": 2.8871644828902174, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.664019584655762, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8609437942504883, "num_tokens": 865984469.0, "step": 22696 }, { "epoch": 2.887291693168808, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.725948333740234, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8753827810287476, "num_tokens": 866024624.0, "step": 22697 }, { "epoch": 2.8874189034473985, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.643288612365723, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8713109493255615, "num_tokens": 866062446.0, "step": 22698 }, { "epoch": 2.887546113725989, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.665445327758789, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8839188814163208, "num_tokens": 866100896.0, "step": 22699 }, { "epoch": 2.8876733240045795, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.719439506530762, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8608136177062988, "num_tokens": 866134668.0, "step": 22700 }, { "epoch": 2.88780053428317, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.689132690429688, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8676179647445679, "num_tokens": 866173834.0, "step": 22701 }, { "epoch": 2.8879277445617606, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.587640762329102, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8590021133422852, "num_tokens": 866211691.0, "step": 22702 }, { "epoch": 2.888054954840351, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.65166187286377, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8740071058273315, "num_tokens": 866240878.0, "step": 22703 }, { "epoch": 2.8881821651189417, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.598414421081543, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8616120219230652, "num_tokens": 866289552.0, "step": 22704 }, { "epoch": 2.888309375397532, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.661019325256348, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8646811842918396, "num_tokens": 866330361.0, "step": 22705 }, { "epoch": 2.8884365856761227, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.638243675231934, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8778529167175293, "num_tokens": 866371557.0, "step": 22706 }, { "epoch": 2.8885637959547132, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.668935775756836, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8747881650924683, "num_tokens": 866410627.0, "step": 22707 }, { "epoch": 2.8886910062333038, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.605828285217285, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8731330633163452, "num_tokens": 866446036.0, "step": 22708 }, { "epoch": 2.8888182165118943, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.613158226013184, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.865725040435791, "num_tokens": 866489655.0, "step": 22709 }, { "epoch": 2.888945426790485, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.651918411254883, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8671688437461853, "num_tokens": 866532750.0, "step": 22710 }, { "epoch": 2.8890726370690754, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.640353202819824, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8804425001144409, "num_tokens": 866570748.0, "step": 22711 }, { "epoch": 2.8891998473476654, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.603289604187012, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8687728643417358, "num_tokens": 866609422.0, "step": 22712 }, { "epoch": 2.8893270576262564, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.555763244628906, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8751651048660278, "num_tokens": 866650865.0, "step": 22713 }, { "epoch": 2.8894542679048465, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.664185523986816, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8730309009552002, "num_tokens": 866687605.0, "step": 22714 }, { "epoch": 2.8895814781834375, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.639827728271484, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.865577220916748, "num_tokens": 866722974.0, "step": 22715 }, { "epoch": 2.8897086884620276, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.64944839477539, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8713893890380859, "num_tokens": 866760694.0, "step": 22716 }, { "epoch": 2.889835898740618, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.623282432556152, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8574481010437012, "num_tokens": 866795102.0, "step": 22717 }, { "epoch": 2.8899631090192086, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.593875885009766, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8877591490745544, "num_tokens": 866831282.0, "step": 22718 }, { "epoch": 2.890090319297799, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.61825942993164, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8663499355316162, "num_tokens": 866872661.0, "step": 22719 }, { "epoch": 2.8902175295763897, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.615689277648926, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8639645576477051, "num_tokens": 866917576.0, "step": 22720 }, { "epoch": 2.89034473985498, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.57180404663086, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.877988874912262, "num_tokens": 866957245.0, "step": 22721 }, { "epoch": 2.8904719501335707, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.649206161499023, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.845933198928833, "num_tokens": 866998695.0, "step": 22722 }, { "epoch": 2.8905991604121613, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.640250205993652, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8614587187767029, "num_tokens": 867033502.0, "step": 22723 }, { "epoch": 2.890726370690752, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.5984525680542, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8560622930526733, "num_tokens": 867076032.0, "step": 22724 }, { "epoch": 2.8908535809693423, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.658209800720215, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8736796379089355, "num_tokens": 867107429.0, "step": 22725 }, { "epoch": 2.890980791247933, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.671976089477539, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8763267397880554, "num_tokens": 867139432.0, "step": 22726 }, { "epoch": 2.8911080015265234, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.576851844787598, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8704261779785156, "num_tokens": 867180588.0, "step": 22727 }, { "epoch": 2.891235211805114, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.672711372375488, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8708645701408386, "num_tokens": 867215170.0, "step": 22728 }, { "epoch": 2.8913624220837044, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.593077659606934, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8879669904708862, "num_tokens": 867254958.0, "step": 22729 }, { "epoch": 2.891489632362295, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.659704208374023, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8725385069847107, "num_tokens": 867291135.0, "step": 22730 }, { "epoch": 2.8916168426408855, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.608149528503418, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.88145911693573, "num_tokens": 867326926.0, "step": 22731 }, { "epoch": 2.891744052919476, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.665751457214355, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8662259578704834, "num_tokens": 867361142.0, "step": 22732 }, { "epoch": 2.8918712631980665, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.550994873046875, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8910450339317322, "num_tokens": 867398344.0, "step": 22733 }, { "epoch": 2.891998473476657, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.684178352355957, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8822847008705139, "num_tokens": 867437642.0, "step": 22734 }, { "epoch": 2.8921256837552476, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.566285133361816, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8639147877693176, "num_tokens": 867471646.0, "step": 22735 }, { "epoch": 2.892252894033838, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.627439498901367, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8739285469055176, "num_tokens": 867509682.0, "step": 22736 }, { "epoch": 2.892380104312428, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.670707702636719, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8700376749038696, "num_tokens": 867544425.0, "step": 22737 }, { "epoch": 2.892507314591019, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.690110206604004, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8665697574615479, "num_tokens": 867584978.0, "step": 22738 }, { "epoch": 2.8926345248696093, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.5993070602417, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8743849396705627, "num_tokens": 867620360.0, "step": 22739 }, { "epoch": 2.8927617351482002, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.666701316833496, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8819767236709595, "num_tokens": 867655246.0, "step": 22740 }, { "epoch": 2.8928889454267903, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.58069896697998, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8657347559928894, "num_tokens": 867692792.0, "step": 22741 }, { "epoch": 2.893016155705381, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.657939910888672, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8765031099319458, "num_tokens": 867732365.0, "step": 22742 }, { "epoch": 2.8931433659839714, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.667906761169434, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.860363245010376, "num_tokens": 867764800.0, "step": 22743 }, { "epoch": 2.893270576262562, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.59585189819336, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8667538166046143, "num_tokens": 867807107.0, "step": 22744 }, { "epoch": 2.8933977865411524, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.638821601867676, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8652362823486328, "num_tokens": 867850225.0, "step": 22745 }, { "epoch": 2.893524996819743, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.589468002319336, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8700897693634033, "num_tokens": 867890457.0, "step": 22746 }, { "epoch": 2.8936522070983335, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.68061637878418, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8800827264785767, "num_tokens": 867925880.0, "step": 22747 }, { "epoch": 2.893779417376924, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.611552238464355, "learning_rate": 1e-06, "loss": 0.3557, "mean_token_accuracy": 0.886441707611084, "num_tokens": 867963923.0, "step": 22748 }, { "epoch": 2.8939066276555145, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.646687507629395, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8774321675300598, "num_tokens": 868000489.0, "step": 22749 }, { "epoch": 2.894033837934105, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.589662551879883, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8766793012619019, "num_tokens": 868034001.0, "step": 22750 }, { "epoch": 2.8941610482126956, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.629209518432617, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8786277174949646, "num_tokens": 868074956.0, "step": 22751 }, { "epoch": 2.894288258491286, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.5909423828125, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8811458349227905, "num_tokens": 868110727.0, "step": 22752 }, { "epoch": 2.8944154687698767, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.620223999023438, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.856479287147522, "num_tokens": 868151480.0, "step": 22753 }, { "epoch": 2.894542679048467, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.617212295532227, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8641597628593445, "num_tokens": 868190620.0, "step": 22754 }, { "epoch": 2.8946698893270577, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.592538833618164, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8716124296188354, "num_tokens": 868230598.0, "step": 22755 }, { "epoch": 2.8947970996056482, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.695352554321289, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8680450916290283, "num_tokens": 868265933.0, "step": 22756 }, { "epoch": 2.8949243098842388, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.60435676574707, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8761978149414062, "num_tokens": 868300282.0, "step": 22757 }, { "epoch": 2.8950515201628293, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.678627967834473, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.888868510723114, "num_tokens": 868331031.0, "step": 22758 }, { "epoch": 2.89517873044142, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.60527229309082, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8764398097991943, "num_tokens": 868368268.0, "step": 22759 }, { "epoch": 2.89530594072001, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.584651947021484, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8589560985565186, "num_tokens": 868407551.0, "step": 22760 }, { "epoch": 2.895433150998601, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.575858116149902, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8714293837547302, "num_tokens": 868449301.0, "step": 22761 }, { "epoch": 2.895560361277191, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.593025207519531, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8716157674789429, "num_tokens": 868482853.0, "step": 22762 }, { "epoch": 2.895687571555782, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.691378593444824, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8699519634246826, "num_tokens": 868523960.0, "step": 22763 }, { "epoch": 2.895814781834372, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.545339584350586, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8657523989677429, "num_tokens": 868566447.0, "step": 22764 }, { "epoch": 2.895941992112963, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.680990219116211, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8833584189414978, "num_tokens": 868608342.0, "step": 22765 }, { "epoch": 2.896069202391553, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.595728874206543, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8853148221969604, "num_tokens": 868647354.0, "step": 22766 }, { "epoch": 2.8961964126701436, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.624059677124023, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.854975700378418, "num_tokens": 868690716.0, "step": 22767 }, { "epoch": 2.896323622948734, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.650582313537598, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8799868822097778, "num_tokens": 868728318.0, "step": 22768 }, { "epoch": 2.8964508332273247, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.697752952575684, "learning_rate": 1e-06, "loss": 0.3502, "mean_token_accuracy": 0.8896808624267578, "num_tokens": 868764323.0, "step": 22769 }, { "epoch": 2.896578043505915, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.668327331542969, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8567423820495605, "num_tokens": 868801371.0, "step": 22770 }, { "epoch": 2.8967052537845057, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.61388111114502, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8785295486450195, "num_tokens": 868846679.0, "step": 22771 }, { "epoch": 2.8968324640630962, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.663363456726074, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8607147932052612, "num_tokens": 868880076.0, "step": 22772 }, { "epoch": 2.8969596743416868, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.595728874206543, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8743318319320679, "num_tokens": 868917086.0, "step": 22773 }, { "epoch": 2.8970868846202773, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.670170783996582, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8808209896087646, "num_tokens": 868954952.0, "step": 22774 }, { "epoch": 2.897214094898868, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.58447551727295, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8695513010025024, "num_tokens": 868990911.0, "step": 22775 }, { "epoch": 2.8973413051774584, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.624038696289062, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8847162127494812, "num_tokens": 869033398.0, "step": 22776 }, { "epoch": 2.897468515456049, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.593940734863281, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8611266613006592, "num_tokens": 869070513.0, "step": 22777 }, { "epoch": 2.8975957257346394, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.628393173217773, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.881749153137207, "num_tokens": 869113494.0, "step": 22778 }, { "epoch": 2.89772293601323, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.672120094299316, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8841419219970703, "num_tokens": 869147433.0, "step": 22779 }, { "epoch": 2.8978501462918205, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.610356330871582, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8698570132255554, "num_tokens": 869186763.0, "step": 22780 }, { "epoch": 2.897977356570411, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.614331245422363, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8868740797042847, "num_tokens": 869222624.0, "step": 22781 }, { "epoch": 2.8981045668490015, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.656088829040527, "learning_rate": 1e-06, "loss": 0.3395, "mean_token_accuracy": 0.8950934410095215, "num_tokens": 869263027.0, "step": 22782 }, { "epoch": 2.898231777127592, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.637965202331543, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8693159818649292, "num_tokens": 869304232.0, "step": 22783 }, { "epoch": 2.8983589874061826, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.618966102600098, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8794960975646973, "num_tokens": 869340141.0, "step": 22784 }, { "epoch": 2.8984861976847727, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.65382194519043, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8651832938194275, "num_tokens": 869382428.0, "step": 22785 }, { "epoch": 2.8986134079633636, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.669997215270996, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8812173008918762, "num_tokens": 869418890.0, "step": 22786 }, { "epoch": 2.8987406182419537, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.632173538208008, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.87525475025177, "num_tokens": 869453704.0, "step": 22787 }, { "epoch": 2.8988678285205447, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.595159530639648, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8711600303649902, "num_tokens": 869491845.0, "step": 22788 }, { "epoch": 2.898995038799135, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.661940574645996, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.885939359664917, "num_tokens": 869523393.0, "step": 22789 }, { "epoch": 2.8991222490777258, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.570066452026367, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8722557425498962, "num_tokens": 869561927.0, "step": 22790 }, { "epoch": 2.899249459356316, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.659354209899902, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.873552680015564, "num_tokens": 869599568.0, "step": 22791 }, { "epoch": 2.8993766696349064, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.628829956054688, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8771301507949829, "num_tokens": 869636329.0, "step": 22792 }, { "epoch": 2.899503879913497, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.625153541564941, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8697788715362549, "num_tokens": 869679720.0, "step": 22793 }, { "epoch": 2.8996310901920874, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.62094497680664, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8768042922019958, "num_tokens": 869719527.0, "step": 22794 }, { "epoch": 2.899758300470678, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.583664894104004, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8792356252670288, "num_tokens": 869755139.0, "step": 22795 }, { "epoch": 2.8998855107492685, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.590198516845703, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8690090179443359, "num_tokens": 869791860.0, "step": 22796 }, { "epoch": 2.900012721027859, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.546609878540039, "learning_rate": 1e-06, "loss": 0.338, "mean_token_accuracy": 0.8956296443939209, "num_tokens": 869827294.0, "step": 22797 }, { "epoch": 2.9001399313064495, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.636512756347656, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8688305616378784, "num_tokens": 869859398.0, "step": 22798 }, { "epoch": 2.90026714158504, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.583379745483398, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.87334805727005, "num_tokens": 869898215.0, "step": 22799 }, { "epoch": 2.9003943518636306, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.618012428283691, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8640976548194885, "num_tokens": 869939787.0, "step": 22800 }, { "epoch": 2.900521562142221, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.546409606933594, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8649085760116577, "num_tokens": 869976790.0, "step": 22801 }, { "epoch": 2.9006487724208116, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.642632484436035, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8784428834915161, "num_tokens": 870013547.0, "step": 22802 }, { "epoch": 2.900775982699402, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.6387357711792, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8854678273200989, "num_tokens": 870049540.0, "step": 22803 }, { "epoch": 2.9009031929779927, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.650018692016602, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8808580636978149, "num_tokens": 870088247.0, "step": 22804 }, { "epoch": 2.9010304032565832, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.586804389953613, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8724143505096436, "num_tokens": 870129112.0, "step": 22805 }, { "epoch": 2.9011576135351738, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.625694274902344, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8735049366950989, "num_tokens": 870162082.0, "step": 22806 }, { "epoch": 2.9012848238137643, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.59316349029541, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8849654197692871, "num_tokens": 870199868.0, "step": 22807 }, { "epoch": 2.901412034092355, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.633611679077148, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8790265321731567, "num_tokens": 870238121.0, "step": 22808 }, { "epoch": 2.9015392443709453, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.596131324768066, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8654462099075317, "num_tokens": 870278008.0, "step": 22809 }, { "epoch": 2.9016664546495354, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.579011917114258, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8719592690467834, "num_tokens": 870317671.0, "step": 22810 }, { "epoch": 2.9017936649281264, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.655561447143555, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8809616565704346, "num_tokens": 870352893.0, "step": 22811 }, { "epoch": 2.9019208752067165, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.635372161865234, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.882566511631012, "num_tokens": 870387513.0, "step": 22812 }, { "epoch": 2.9020480854853075, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.592428207397461, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.880620539188385, "num_tokens": 870428649.0, "step": 22813 }, { "epoch": 2.9021752957638975, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.640483856201172, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8878002762794495, "num_tokens": 870468329.0, "step": 22814 }, { "epoch": 2.902302506042488, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.67074203491211, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8734260201454163, "num_tokens": 870509273.0, "step": 22815 }, { "epoch": 2.9024297163210786, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.65117073059082, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8690841197967529, "num_tokens": 870544206.0, "step": 22816 }, { "epoch": 2.902556926599669, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.601195335388184, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8911463022232056, "num_tokens": 870578708.0, "step": 22817 }, { "epoch": 2.9026841368782597, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.667437553405762, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8814216256141663, "num_tokens": 870620599.0, "step": 22818 }, { "epoch": 2.90281134715685, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.615787506103516, "learning_rate": 1e-06, "loss": 0.3494, "mean_token_accuracy": 0.888304591178894, "num_tokens": 870657828.0, "step": 22819 }, { "epoch": 2.9029385574354407, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.608015060424805, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8690484762191772, "num_tokens": 870694786.0, "step": 22820 }, { "epoch": 2.9030657677140312, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.520060539245605, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8915455937385559, "num_tokens": 870733850.0, "step": 22821 }, { "epoch": 2.9031929779926218, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.650382995605469, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8747125864028931, "num_tokens": 870771556.0, "step": 22822 }, { "epoch": 2.9033201882712123, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.607784271240234, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8859927654266357, "num_tokens": 870809335.0, "step": 22823 }, { "epoch": 2.903447398549803, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.612564086914062, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8704327344894409, "num_tokens": 870844239.0, "step": 22824 }, { "epoch": 2.9035746088283934, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.671918869018555, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8737424612045288, "num_tokens": 870879540.0, "step": 22825 }, { "epoch": 2.903701819106984, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.585624694824219, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8745141625404358, "num_tokens": 870917219.0, "step": 22826 }, { "epoch": 2.9038290293855744, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.626870155334473, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8756994009017944, "num_tokens": 870952332.0, "step": 22827 }, { "epoch": 2.903956239664165, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.630509376525879, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8634828329086304, "num_tokens": 870991609.0, "step": 22828 }, { "epoch": 2.9040834499427555, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.62336254119873, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8782296180725098, "num_tokens": 871027456.0, "step": 22829 }, { "epoch": 2.904210660221346, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.651301383972168, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8655815124511719, "num_tokens": 871071833.0, "step": 22830 }, { "epoch": 2.9043378704999365, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.654350280761719, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8737634420394897, "num_tokens": 871105446.0, "step": 22831 }, { "epoch": 2.904465080778527, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.621315956115723, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8737483024597168, "num_tokens": 871141408.0, "step": 22832 }, { "epoch": 2.9045922910571176, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.741157531738281, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8779083490371704, "num_tokens": 871173296.0, "step": 22833 }, { "epoch": 2.904719501335708, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.593327522277832, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8714607357978821, "num_tokens": 871210711.0, "step": 22834 }, { "epoch": 2.904846711614298, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.668289184570312, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8719771504402161, "num_tokens": 871251179.0, "step": 22835 }, { "epoch": 2.904973921892889, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.640524864196777, "learning_rate": 1e-06, "loss": 0.3393, "mean_token_accuracy": 0.8934904336929321, "num_tokens": 871295758.0, "step": 22836 }, { "epoch": 2.9051011321714793, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.656593322753906, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8597464561462402, "num_tokens": 871331145.0, "step": 22837 }, { "epoch": 2.9052283424500702, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.657865524291992, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8505179286003113, "num_tokens": 871372819.0, "step": 22838 }, { "epoch": 2.9053555527286603, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.66016674041748, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8693345785140991, "num_tokens": 871414223.0, "step": 22839 }, { "epoch": 2.905482763007251, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.641203880310059, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8800739049911499, "num_tokens": 871455013.0, "step": 22840 }, { "epoch": 2.9056099732858414, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.636354446411133, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8734731674194336, "num_tokens": 871495462.0, "step": 22841 }, { "epoch": 2.905737183564432, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.652113914489746, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8773683309555054, "num_tokens": 871534999.0, "step": 22842 }, { "epoch": 2.9058643938430224, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.652612686157227, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8749665021896362, "num_tokens": 871569816.0, "step": 22843 }, { "epoch": 2.905991604121613, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.603180885314941, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.877275288105011, "num_tokens": 871610441.0, "step": 22844 }, { "epoch": 2.9061188144002035, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.601921081542969, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8609751462936401, "num_tokens": 871649005.0, "step": 22845 }, { "epoch": 2.906246024678794, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.556553840637207, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8700777292251587, "num_tokens": 871689263.0, "step": 22846 }, { "epoch": 2.9063732349573845, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.565987586975098, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8830448389053345, "num_tokens": 871725440.0, "step": 22847 }, { "epoch": 2.906500445235975, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.631163597106934, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8720924258232117, "num_tokens": 871767839.0, "step": 22848 }, { "epoch": 2.9066276555145656, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.608495712280273, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8789809942245483, "num_tokens": 871801516.0, "step": 22849 }, { "epoch": 2.906754865793156, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.56478500366211, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.870262086391449, "num_tokens": 871840359.0, "step": 22850 }, { "epoch": 2.9068820760717466, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.645608901977539, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8789423108100891, "num_tokens": 871879396.0, "step": 22851 }, { "epoch": 2.907009286350337, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.626209259033203, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8675999641418457, "num_tokens": 871915043.0, "step": 22852 }, { "epoch": 2.9071364966289277, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.619769096374512, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.869814932346344, "num_tokens": 871953391.0, "step": 22853 }, { "epoch": 2.9072637069075182, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.633805274963379, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8779854774475098, "num_tokens": 871989879.0, "step": 22854 }, { "epoch": 2.9073909171861088, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.63833999633789, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8778840899467468, "num_tokens": 872022428.0, "step": 22855 }, { "epoch": 2.9075181274646993, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.642845153808594, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8798089623451233, "num_tokens": 872060644.0, "step": 22856 }, { "epoch": 2.90764533774329, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.574928283691406, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8724507093429565, "num_tokens": 872098230.0, "step": 22857 }, { "epoch": 2.90777254802188, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.62346076965332, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8634659051895142, "num_tokens": 872131217.0, "step": 22858 }, { "epoch": 2.907899758300471, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.536629676818848, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8841530084609985, "num_tokens": 872163912.0, "step": 22859 }, { "epoch": 2.908026968579061, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.594525337219238, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.851012647151947, "num_tokens": 872200694.0, "step": 22860 }, { "epoch": 2.908154178857652, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.591071128845215, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8754386901855469, "num_tokens": 872241156.0, "step": 22861 }, { "epoch": 2.908281389136242, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.625930786132812, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8771550059318542, "num_tokens": 872273048.0, "step": 22862 }, { "epoch": 2.908408599414833, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.57058048248291, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8779232501983643, "num_tokens": 872312165.0, "step": 22863 }, { "epoch": 2.908535809693423, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.578262329101562, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8767815232276917, "num_tokens": 872352740.0, "step": 22864 }, { "epoch": 2.9086630199720136, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.600781440734863, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.860816240310669, "num_tokens": 872397800.0, "step": 22865 }, { "epoch": 2.908790230250604, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.563875198364258, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8856629729270935, "num_tokens": 872429307.0, "step": 22866 }, { "epoch": 2.9089174405291947, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.666949272155762, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8717240691184998, "num_tokens": 872465897.0, "step": 22867 }, { "epoch": 2.909044650807785, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.642044067382812, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.865759015083313, "num_tokens": 872503252.0, "step": 22868 }, { "epoch": 2.9091718610863757, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.618144035339355, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8773236870765686, "num_tokens": 872546276.0, "step": 22869 }, { "epoch": 2.9092990713649662, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.617314338684082, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8727059960365295, "num_tokens": 872585690.0, "step": 22870 }, { "epoch": 2.9094262816435568, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.650799751281738, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8643333911895752, "num_tokens": 872621718.0, "step": 22871 }, { "epoch": 2.9095534919221473, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.573221206665039, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8785210847854614, "num_tokens": 872659349.0, "step": 22872 }, { "epoch": 2.909680702200738, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.576969146728516, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8705871105194092, "num_tokens": 872704468.0, "step": 22873 }, { "epoch": 2.9098079124793284, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.611425399780273, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8756119608879089, "num_tokens": 872740003.0, "step": 22874 }, { "epoch": 2.909935122757919, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.580816268920898, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8528590798377991, "num_tokens": 872780362.0, "step": 22875 }, { "epoch": 2.9100623330365094, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.675146102905273, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8822081089019775, "num_tokens": 872823719.0, "step": 22876 }, { "epoch": 2.9101895433151, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.592596054077148, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.878716766834259, "num_tokens": 872856276.0, "step": 22877 }, { "epoch": 2.9103167535936905, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.66365909576416, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8681423664093018, "num_tokens": 872893688.0, "step": 22878 }, { "epoch": 2.910443963872281, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.637349128723145, "learning_rate": 1e-06, "loss": 0.3411, "mean_token_accuracy": 0.8926136493682861, "num_tokens": 872931169.0, "step": 22879 }, { "epoch": 2.9105711741508715, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.62349796295166, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8698402047157288, "num_tokens": 872970836.0, "step": 22880 }, { "epoch": 2.910698384429462, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.625914573669434, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8876891136169434, "num_tokens": 873010879.0, "step": 22881 }, { "epoch": 2.9108255947080526, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.665617942810059, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8763450980186462, "num_tokens": 873058018.0, "step": 22882 }, { "epoch": 2.9109528049866427, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.618620872497559, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8633520603179932, "num_tokens": 873096259.0, "step": 22883 }, { "epoch": 2.9110800152652336, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.703774452209473, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8646724820137024, "num_tokens": 873131955.0, "step": 22884 }, { "epoch": 2.9112072255438237, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.630407333374023, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8689992427825928, "num_tokens": 873171024.0, "step": 22885 }, { "epoch": 2.9113344358224147, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.590583801269531, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8777768611907959, "num_tokens": 873210568.0, "step": 22886 }, { "epoch": 2.9114616461010048, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.645659446716309, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8770300149917603, "num_tokens": 873250025.0, "step": 22887 }, { "epoch": 2.9115888563795957, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.582024574279785, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8662418723106384, "num_tokens": 873285284.0, "step": 22888 }, { "epoch": 2.911716066658186, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.644949913024902, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8771425485610962, "num_tokens": 873320102.0, "step": 22889 }, { "epoch": 2.9118432769367764, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.594108581542969, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8791033625602722, "num_tokens": 873354668.0, "step": 22890 }, { "epoch": 2.911970487215367, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.70124626159668, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8720664978027344, "num_tokens": 873387415.0, "step": 22891 }, { "epoch": 2.9120976974939574, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.527321815490723, "learning_rate": 1e-06, "loss": 0.3491, "mean_token_accuracy": 0.8915428519248962, "num_tokens": 873428679.0, "step": 22892 }, { "epoch": 2.912224907772548, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.598082542419434, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.876136064529419, "num_tokens": 873466973.0, "step": 22893 }, { "epoch": 2.9123521180511385, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.575221061706543, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.8891407251358032, "num_tokens": 873510957.0, "step": 22894 }, { "epoch": 2.912479328329729, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.654980659484863, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8549169898033142, "num_tokens": 873546456.0, "step": 22895 }, { "epoch": 2.9126065386083195, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.505209922790527, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8609907627105713, "num_tokens": 873591972.0, "step": 22896 }, { "epoch": 2.91273374888691, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.643847465515137, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8730371594429016, "num_tokens": 873629790.0, "step": 22897 }, { "epoch": 2.9128609591655006, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.6026029586792, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8748947978019714, "num_tokens": 873667419.0, "step": 22898 }, { "epoch": 2.912988169444091, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.628581047058105, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8837813138961792, "num_tokens": 873705834.0, "step": 22899 }, { "epoch": 2.9131153797226816, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.643180847167969, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8657126426696777, "num_tokens": 873748859.0, "step": 22900 }, { "epoch": 2.913242590001272, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.581243515014648, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.86601322889328, "num_tokens": 873788068.0, "step": 22901 }, { "epoch": 2.9133698002798627, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.646300315856934, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8825796246528625, "num_tokens": 873827820.0, "step": 22902 }, { "epoch": 2.9134970105584532, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.612763404846191, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8708444833755493, "num_tokens": 873866271.0, "step": 22903 }, { "epoch": 2.9136242208370438, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.613747596740723, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8712500929832458, "num_tokens": 873907409.0, "step": 22904 }, { "epoch": 2.9137514311156343, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.698079109191895, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.879687488079071, "num_tokens": 873947940.0, "step": 22905 }, { "epoch": 2.913878641394225, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.602879524230957, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.859896183013916, "num_tokens": 873990062.0, "step": 22906 }, { "epoch": 2.9140058516728153, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.70036506652832, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8856209516525269, "num_tokens": 874026660.0, "step": 22907 }, { "epoch": 2.9141330619514054, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.600914001464844, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8811547756195068, "num_tokens": 874064935.0, "step": 22908 }, { "epoch": 2.9142602722299964, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68533992767334, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.873909592628479, "num_tokens": 874102807.0, "step": 22909 }, { "epoch": 2.9143874825085865, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.742044448852539, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.876818835735321, "num_tokens": 874139688.0, "step": 22910 }, { "epoch": 2.9145146927871775, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.551597595214844, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8794520497322083, "num_tokens": 874184505.0, "step": 22911 }, { "epoch": 2.9146419030657675, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.689566612243652, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8743126392364502, "num_tokens": 874223210.0, "step": 22912 }, { "epoch": 2.914769113344358, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.646587371826172, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.861634373664856, "num_tokens": 874263939.0, "step": 22913 }, { "epoch": 2.9148963236229486, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.642021179199219, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8824460506439209, "num_tokens": 874303946.0, "step": 22914 }, { "epoch": 2.915023533901539, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.571151733398438, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8760921955108643, "num_tokens": 874343701.0, "step": 22915 }, { "epoch": 2.9151507441801296, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.639397621154785, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8775560855865479, "num_tokens": 874380425.0, "step": 22916 }, { "epoch": 2.91527795445872, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.640229225158691, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8748519420623779, "num_tokens": 874417643.0, "step": 22917 }, { "epoch": 2.9154051647373107, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.582573890686035, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8673830032348633, "num_tokens": 874463939.0, "step": 22918 }, { "epoch": 2.9155323750159012, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.627039909362793, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8711284399032593, "num_tokens": 874503156.0, "step": 22919 }, { "epoch": 2.9156595852944918, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.673569679260254, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8645773530006409, "num_tokens": 874546754.0, "step": 22920 }, { "epoch": 2.9157867955730823, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.640887260437012, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8775366544723511, "num_tokens": 874590781.0, "step": 22921 }, { "epoch": 2.915914005851673, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.644096374511719, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8835229277610779, "num_tokens": 874629592.0, "step": 22922 }, { "epoch": 2.9160412161302633, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.645492553710938, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8820000886917114, "num_tokens": 874667372.0, "step": 22923 }, { "epoch": 2.916168426408854, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.725372314453125, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8720382452011108, "num_tokens": 874704687.0, "step": 22924 }, { "epoch": 2.9162956366874444, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.560669898986816, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8729950189590454, "num_tokens": 874740471.0, "step": 22925 }, { "epoch": 2.916422846966035, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.636722564697266, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8800712823867798, "num_tokens": 874780747.0, "step": 22926 }, { "epoch": 2.9165500572446255, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.669530868530273, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8792313933372498, "num_tokens": 874814204.0, "step": 22927 }, { "epoch": 2.916677267523216, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.646674156188965, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.873563826084137, "num_tokens": 874851872.0, "step": 22928 }, { "epoch": 2.9168044778018065, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.63230037689209, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8596974015235901, "num_tokens": 874884215.0, "step": 22929 }, { "epoch": 2.916931688080397, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.652033805847168, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8724967837333679, "num_tokens": 874919047.0, "step": 22930 }, { "epoch": 2.917058898358987, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.654705047607422, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.867872953414917, "num_tokens": 874956469.0, "step": 22931 }, { "epoch": 2.917186108637578, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.574353218078613, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8862224221229553, "num_tokens": 874994633.0, "step": 22932 }, { "epoch": 2.917313318916168, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.72330379486084, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8878109455108643, "num_tokens": 875028530.0, "step": 22933 }, { "epoch": 2.917440529194759, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.604909896850586, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8667250871658325, "num_tokens": 875059785.0, "step": 22934 }, { "epoch": 2.9175677394733492, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.891313552856445, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8727463483810425, "num_tokens": 875096057.0, "step": 22935 }, { "epoch": 2.91769494975194, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.631110191345215, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8802909851074219, "num_tokens": 875141988.0, "step": 22936 }, { "epoch": 2.9178221600305303, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.597975730895996, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8796223402023315, "num_tokens": 875176973.0, "step": 22937 }, { "epoch": 2.917949370309121, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.646589279174805, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8833800554275513, "num_tokens": 875209271.0, "step": 22938 }, { "epoch": 2.9180765805877114, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.601845741271973, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8797414302825928, "num_tokens": 875247460.0, "step": 22939 }, { "epoch": 2.918203790866302, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.58991527557373, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8734346628189087, "num_tokens": 875286013.0, "step": 22940 }, { "epoch": 2.9183310011448924, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.591423988342285, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.871436357498169, "num_tokens": 875324547.0, "step": 22941 }, { "epoch": 2.918458211423483, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.625767707824707, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8661297559738159, "num_tokens": 875366655.0, "step": 22942 }, { "epoch": 2.9185854217020735, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.652975082397461, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8692107200622559, "num_tokens": 875408571.0, "step": 22943 }, { "epoch": 2.918712631980664, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68331527709961, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8687893152236938, "num_tokens": 875450730.0, "step": 22944 }, { "epoch": 2.9188398422592545, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.681134223937988, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8783425092697144, "num_tokens": 875486850.0, "step": 22945 }, { "epoch": 2.918967052537845, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.626789093017578, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8688232898712158, "num_tokens": 875528653.0, "step": 22946 }, { "epoch": 2.9190942628164356, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.66460132598877, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8844919204711914, "num_tokens": 875564155.0, "step": 22947 }, { "epoch": 2.919221473095026, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.630850791931152, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8738350868225098, "num_tokens": 875598548.0, "step": 22948 }, { "epoch": 2.9193486833736166, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.717720985412598, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.889756441116333, "num_tokens": 875634587.0, "step": 22949 }, { "epoch": 2.919475893652207, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.659414291381836, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.873790442943573, "num_tokens": 875671956.0, "step": 22950 }, { "epoch": 2.9196031039307977, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.619122505187988, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8769310712814331, "num_tokens": 875706327.0, "step": 22951 }, { "epoch": 2.9197303142093882, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.621103286743164, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8811677694320679, "num_tokens": 875749320.0, "step": 22952 }, { "epoch": 2.9198575244879788, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.694475173950195, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8841131925582886, "num_tokens": 875783871.0, "step": 22953 }, { "epoch": 2.9199847347665693, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.641626358032227, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8738709688186646, "num_tokens": 875821716.0, "step": 22954 }, { "epoch": 2.92011194504516, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.669222831726074, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8773206472396851, "num_tokens": 875854339.0, "step": 22955 }, { "epoch": 2.92023915532375, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.694534301757812, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.875357985496521, "num_tokens": 875889456.0, "step": 22956 }, { "epoch": 2.920366365602341, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.61396598815918, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8663066625595093, "num_tokens": 875926659.0, "step": 22957 }, { "epoch": 2.920493575880931, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.735688209533691, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8819805383682251, "num_tokens": 875972766.0, "step": 22958 }, { "epoch": 2.920620786159522, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.623523712158203, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8701334595680237, "num_tokens": 876011824.0, "step": 22959 }, { "epoch": 2.920747996438112, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.72235107421875, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8576648235321045, "num_tokens": 876050353.0, "step": 22960 }, { "epoch": 2.920875206716703, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.669330596923828, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.860617995262146, "num_tokens": 876089358.0, "step": 22961 }, { "epoch": 2.921002416995293, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.666516304016113, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8611794710159302, "num_tokens": 876126287.0, "step": 22962 }, { "epoch": 2.9211296272738836, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.61160659790039, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8829147815704346, "num_tokens": 876165157.0, "step": 22963 }, { "epoch": 2.921256837552474, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.736632347106934, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8753308057785034, "num_tokens": 876201391.0, "step": 22964 }, { "epoch": 2.9213840478310646, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.58227252960205, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8807312846183777, "num_tokens": 876243004.0, "step": 22965 }, { "epoch": 2.921511258109655, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.668878555297852, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8770574927330017, "num_tokens": 876279245.0, "step": 22966 }, { "epoch": 2.9216384683882457, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.61637020111084, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.87379390001297, "num_tokens": 876312145.0, "step": 22967 }, { "epoch": 2.9217656786668362, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.654879570007324, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8758280277252197, "num_tokens": 876351354.0, "step": 22968 }, { "epoch": 2.9218928889454268, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.598687171936035, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.863579273223877, "num_tokens": 876388621.0, "step": 22969 }, { "epoch": 2.9220200992240173, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.698634147644043, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8688584566116333, "num_tokens": 876418377.0, "step": 22970 }, { "epoch": 2.922147309502608, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.697815895080566, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8767097592353821, "num_tokens": 876453351.0, "step": 22971 }, { "epoch": 2.9222745197811983, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.623076438903809, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8810260891914368, "num_tokens": 876484830.0, "step": 22972 }, { "epoch": 2.922401730059789, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.658934593200684, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8849459886550903, "num_tokens": 876524318.0, "step": 22973 }, { "epoch": 2.9225289403383794, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.644171714782715, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8719478249549866, "num_tokens": 876562421.0, "step": 22974 }, { "epoch": 2.92265615061697, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.66664981842041, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8681325912475586, "num_tokens": 876605681.0, "step": 22975 }, { "epoch": 2.9227833608955605, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.660000801086426, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8656590580940247, "num_tokens": 876647371.0, "step": 22976 }, { "epoch": 2.922910571174151, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.684549331665039, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8796653747558594, "num_tokens": 876685563.0, "step": 22977 }, { "epoch": 2.9230377814527415, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.706958770751953, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8705563545227051, "num_tokens": 876718404.0, "step": 22978 }, { "epoch": 2.923164991731332, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.65987777709961, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8778085112571716, "num_tokens": 876751888.0, "step": 22979 }, { "epoch": 2.9232922020099226, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.706974029541016, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8795945644378662, "num_tokens": 876786813.0, "step": 22980 }, { "epoch": 2.9234194122885127, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.703228950500488, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8835924863815308, "num_tokens": 876824511.0, "step": 22981 }, { "epoch": 2.9235466225671036, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.628336906433105, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8776061534881592, "num_tokens": 876856409.0, "step": 22982 }, { "epoch": 2.9236738328456937, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.699230194091797, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8603941202163696, "num_tokens": 876889204.0, "step": 22983 }, { "epoch": 2.9238010431242847, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.650083541870117, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.875075101852417, "num_tokens": 876927185.0, "step": 22984 }, { "epoch": 2.9239282534028748, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.679555892944336, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8719322681427002, "num_tokens": 876965744.0, "step": 22985 }, { "epoch": 2.9240554636814657, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.718340873718262, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.880699098110199, "num_tokens": 876997107.0, "step": 22986 }, { "epoch": 2.924182673960056, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.652132987976074, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8650150895118713, "num_tokens": 877037894.0, "step": 22987 }, { "epoch": 2.9243098842386464, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.68113899230957, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8616353273391724, "num_tokens": 877078724.0, "step": 22988 }, { "epoch": 2.924437094517237, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.66860294342041, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8746269941329956, "num_tokens": 877119685.0, "step": 22989 }, { "epoch": 2.9245643047958274, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.661425590515137, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8783862590789795, "num_tokens": 877157055.0, "step": 22990 }, { "epoch": 2.924691515074418, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.702967643737793, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8804081678390503, "num_tokens": 877188569.0, "step": 22991 }, { "epoch": 2.9248187253530085, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.659613609313965, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8670045137405396, "num_tokens": 877222745.0, "step": 22992 }, { "epoch": 2.924945935631599, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.635141372680664, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8597487211227417, "num_tokens": 877262510.0, "step": 22993 }, { "epoch": 2.9250731459101895, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.649984359741211, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8681225776672363, "num_tokens": 877300164.0, "step": 22994 }, { "epoch": 2.92520035618878, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.665190696716309, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8843042254447937, "num_tokens": 877337187.0, "step": 22995 }, { "epoch": 2.9253275664673706, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.639931678771973, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.855305552482605, "num_tokens": 877375902.0, "step": 22996 }, { "epoch": 2.925454776745961, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.665966987609863, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8728600740432739, "num_tokens": 877414406.0, "step": 22997 }, { "epoch": 2.9255819870245516, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.582724571228027, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8808144330978394, "num_tokens": 877448977.0, "step": 22998 }, { "epoch": 2.925709197303142, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.649765968322754, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8815685510635376, "num_tokens": 877486766.0, "step": 22999 }, { "epoch": 2.9258364075817327, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.55282211303711, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.887439489364624, "num_tokens": 877519320.0, "step": 23000 }, { "epoch": 2.925963617860323, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.718642234802246, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.850409984588623, "num_tokens": 877557913.0, "step": 23001 }, { "epoch": 2.9260908281389137, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.659027099609375, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8748951554298401, "num_tokens": 877591197.0, "step": 23002 }, { "epoch": 2.9262180384175043, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.58005428314209, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.8905122876167297, "num_tokens": 877625018.0, "step": 23003 }, { "epoch": 2.926345248696095, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.689712524414062, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8795938491821289, "num_tokens": 877662433.0, "step": 23004 }, { "epoch": 2.9264724589746853, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.736465454101562, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8863255381584167, "num_tokens": 877703576.0, "step": 23005 }, { "epoch": 2.9265996692532754, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.664055824279785, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8703580498695374, "num_tokens": 877734025.0, "step": 23006 }, { "epoch": 2.9267268795318664, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.659113883972168, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8835549354553223, "num_tokens": 877771942.0, "step": 23007 }, { "epoch": 2.9268540898104565, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.650260925292969, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8772757649421692, "num_tokens": 877815662.0, "step": 23008 }, { "epoch": 2.9269813000890474, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.631175994873047, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8594033122062683, "num_tokens": 877855398.0, "step": 23009 }, { "epoch": 2.9271085103676375, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.66528606414795, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.868789553642273, "num_tokens": 877898544.0, "step": 23010 }, { "epoch": 2.927235720646228, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.62806510925293, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8733473420143127, "num_tokens": 877937914.0, "step": 23011 }, { "epoch": 2.9273629309248186, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.674766540527344, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8577233552932739, "num_tokens": 877979266.0, "step": 23012 }, { "epoch": 2.927490141203409, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.664645195007324, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8789423704147339, "num_tokens": 878019762.0, "step": 23013 }, { "epoch": 2.9276173514819996, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.6026029586792, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8695259690284729, "num_tokens": 878062174.0, "step": 23014 }, { "epoch": 2.92774456176059, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.699092864990234, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8692213296890259, "num_tokens": 878098197.0, "step": 23015 }, { "epoch": 2.9278717720391807, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.705907821655273, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.880187451839447, "num_tokens": 878130478.0, "step": 23016 }, { "epoch": 2.9279989823177712, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.633188247680664, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8566713333129883, "num_tokens": 878175670.0, "step": 23017 }, { "epoch": 2.9281261925963618, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.642629623413086, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8627679944038391, "num_tokens": 878216415.0, "step": 23018 }, { "epoch": 2.9282534028749523, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.671045303344727, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8734139800071716, "num_tokens": 878249582.0, "step": 23019 }, { "epoch": 2.928380613153543, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.627415657043457, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8845807313919067, "num_tokens": 878290859.0, "step": 23020 }, { "epoch": 2.9285078234321333, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.614259719848633, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8902308344841003, "num_tokens": 878322463.0, "step": 23021 }, { "epoch": 2.928635033710724, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.743646621704102, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8844747543334961, "num_tokens": 878365754.0, "step": 23022 }, { "epoch": 2.9287622439893144, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.607101440429688, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.887069821357727, "num_tokens": 878410325.0, "step": 23023 }, { "epoch": 2.928889454267905, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.684551239013672, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8791323900222778, "num_tokens": 878451604.0, "step": 23024 }, { "epoch": 2.9290166645464955, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.690709114074707, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8775495886802673, "num_tokens": 878485174.0, "step": 23025 }, { "epoch": 2.929143874825086, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.718951225280762, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8667572140693665, "num_tokens": 878525867.0, "step": 23026 }, { "epoch": 2.9292710851036765, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.609728813171387, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.862666130065918, "num_tokens": 878561327.0, "step": 23027 }, { "epoch": 2.929398295382267, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.66297721862793, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8807762265205383, "num_tokens": 878603045.0, "step": 23028 }, { "epoch": 2.929525505660857, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.598764419555664, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8805333375930786, "num_tokens": 878645051.0, "step": 23029 }, { "epoch": 2.929652715939448, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.665074348449707, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8568981885910034, "num_tokens": 878682003.0, "step": 23030 }, { "epoch": 2.929779926218038, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.704378128051758, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8801929950714111, "num_tokens": 878720509.0, "step": 23031 }, { "epoch": 2.929907136496629, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.606767654418945, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8801413178443909, "num_tokens": 878757950.0, "step": 23032 }, { "epoch": 2.9300343467752192, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.63652515411377, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8678682446479797, "num_tokens": 878801477.0, "step": 23033 }, { "epoch": 2.93016155705381, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.696852684020996, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.867024838924408, "num_tokens": 878838802.0, "step": 23034 }, { "epoch": 2.9302887673324003, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.649547576904297, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8698720932006836, "num_tokens": 878875566.0, "step": 23035 }, { "epoch": 2.930415977610991, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.641223907470703, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8673522472381592, "num_tokens": 878913463.0, "step": 23036 }, { "epoch": 2.9305431878895813, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.673026084899902, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.862947940826416, "num_tokens": 878950769.0, "step": 23037 }, { "epoch": 2.930670398168172, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.64888858795166, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8849900364875793, "num_tokens": 878989041.0, "step": 23038 }, { "epoch": 2.9307976084467624, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.745776176452637, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8816369771957397, "num_tokens": 879020195.0, "step": 23039 }, { "epoch": 2.930924818725353, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.699455261230469, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8717042207717896, "num_tokens": 879057818.0, "step": 23040 }, { "epoch": 2.9310520290039435, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.62890338897705, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8712584972381592, "num_tokens": 879094747.0, "step": 23041 }, { "epoch": 2.931179239282534, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.708712577819824, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8660154938697815, "num_tokens": 879131127.0, "step": 23042 }, { "epoch": 2.9313064495611245, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.662678718566895, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.874999463558197, "num_tokens": 879168541.0, "step": 23043 }, { "epoch": 2.931433659839715, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.69573974609375, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.871393084526062, "num_tokens": 879199409.0, "step": 23044 }, { "epoch": 2.9315608701183056, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.710965156555176, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8875821828842163, "num_tokens": 879236447.0, "step": 23045 }, { "epoch": 2.931688080396896, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.665675163269043, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8749063014984131, "num_tokens": 879275596.0, "step": 23046 }, { "epoch": 2.9318152906754866, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.697300910949707, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8731312155723572, "num_tokens": 879315620.0, "step": 23047 }, { "epoch": 2.931942500954077, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.754533767700195, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8763290643692017, "num_tokens": 879356220.0, "step": 23048 }, { "epoch": 2.9320697112326677, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.699884414672852, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8595378994941711, "num_tokens": 879397544.0, "step": 23049 }, { "epoch": 2.932196921511258, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719897270202637, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8843695521354675, "num_tokens": 879438165.0, "step": 23050 }, { "epoch": 2.9323241317898487, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.682543754577637, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.875468909740448, "num_tokens": 879473601.0, "step": 23051 }, { "epoch": 2.9324513420684393, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.784856796264648, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.867477297782898, "num_tokens": 879510998.0, "step": 23052 }, { "epoch": 2.93257855234703, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.648944854736328, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8789968490600586, "num_tokens": 879551469.0, "step": 23053 }, { "epoch": 2.93270576262562, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.788225173950195, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8675000071525574, "num_tokens": 879585620.0, "step": 23054 }, { "epoch": 2.932832972904211, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.691424369812012, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8839375972747803, "num_tokens": 879621749.0, "step": 23055 }, { "epoch": 2.932960183182801, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.711068153381348, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8778713941574097, "num_tokens": 879664529.0, "step": 23056 }, { "epoch": 2.933087393461392, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.70459270477295, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8796917200088501, "num_tokens": 879696293.0, "step": 23057 }, { "epoch": 2.933214603739982, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.736477851867676, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8788174390792847, "num_tokens": 879733562.0, "step": 23058 }, { "epoch": 2.933341814018573, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.59961986541748, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8724780082702637, "num_tokens": 879771740.0, "step": 23059 }, { "epoch": 2.933469024297163, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.844032287597656, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8700425624847412, "num_tokens": 879805578.0, "step": 23060 }, { "epoch": 2.9335962345757536, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.652429580688477, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.871538519859314, "num_tokens": 879845175.0, "step": 23061 }, { "epoch": 2.933723444854344, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.723854064941406, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8855743408203125, "num_tokens": 879878975.0, "step": 23062 }, { "epoch": 2.9338506551329346, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.66266918182373, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8657255172729492, "num_tokens": 879919373.0, "step": 23063 }, { "epoch": 2.933977865411525, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.827972412109375, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8547969460487366, "num_tokens": 879953821.0, "step": 23064 }, { "epoch": 2.9341050756901157, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.674883842468262, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.866229236125946, "num_tokens": 879993505.0, "step": 23065 }, { "epoch": 2.934232285968706, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.727829933166504, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8793845176696777, "num_tokens": 880030124.0, "step": 23066 }, { "epoch": 2.9343594962472968, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.786351203918457, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8521210551261902, "num_tokens": 880059480.0, "step": 23067 }, { "epoch": 2.9344867065258873, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.777379035949707, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8750548362731934, "num_tokens": 880093402.0, "step": 23068 }, { "epoch": 2.934613916804478, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.720893859863281, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8726145029067993, "num_tokens": 880127874.0, "step": 23069 }, { "epoch": 2.9347411270830683, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.76541805267334, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8839825987815857, "num_tokens": 880165755.0, "step": 23070 }, { "epoch": 2.934868337361659, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.708579063415527, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8687988519668579, "num_tokens": 880203931.0, "step": 23071 }, { "epoch": 2.9349955476402494, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.680278778076172, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8767358064651489, "num_tokens": 880242844.0, "step": 23072 }, { "epoch": 2.93512275791884, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.763466835021973, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8596530556678772, "num_tokens": 880280068.0, "step": 23073 }, { "epoch": 2.9352499681974304, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.692977905273438, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8800907135009766, "num_tokens": 880314044.0, "step": 23074 }, { "epoch": 2.935377178476021, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.714181900024414, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8703843355178833, "num_tokens": 880355077.0, "step": 23075 }, { "epoch": 2.9355043887546115, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.675117492675781, "learning_rate": 1e-06, "loss": 0.3502, "mean_token_accuracy": 0.8895542621612549, "num_tokens": 880388906.0, "step": 23076 }, { "epoch": 2.935631599033202, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.736839294433594, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8855377435684204, "num_tokens": 880426834.0, "step": 23077 }, { "epoch": 2.9357588093117926, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.641371726989746, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.872582197189331, "num_tokens": 880463477.0, "step": 23078 }, { "epoch": 2.9358860195903826, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.819860458374023, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8572871088981628, "num_tokens": 880500884.0, "step": 23079 }, { "epoch": 2.9360132298689736, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.733388900756836, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8679726123809814, "num_tokens": 880539724.0, "step": 23080 }, { "epoch": 2.9361404401475637, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68436050415039, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.867664098739624, "num_tokens": 880577820.0, "step": 23081 }, { "epoch": 2.9362676504261547, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.672924041748047, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8780786395072937, "num_tokens": 880612055.0, "step": 23082 }, { "epoch": 2.9363948607047448, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.679570198059082, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8815718293190002, "num_tokens": 880654907.0, "step": 23083 }, { "epoch": 2.9365220709833357, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.749198913574219, "learning_rate": 1e-06, "loss": 0.3443, "mean_token_accuracy": 0.8902508020401001, "num_tokens": 880692418.0, "step": 23084 }, { "epoch": 2.936649281261926, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.762046813964844, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.860541582107544, "num_tokens": 880727742.0, "step": 23085 }, { "epoch": 2.9367764915405163, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.633126258850098, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8859645128250122, "num_tokens": 880766673.0, "step": 23086 }, { "epoch": 2.936903701819107, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.676712036132812, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8666121363639832, "num_tokens": 880810390.0, "step": 23087 }, { "epoch": 2.9370309120976974, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.738314628601074, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8690195083618164, "num_tokens": 880846920.0, "step": 23088 }, { "epoch": 2.937158122376288, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.639547348022461, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8809084296226501, "num_tokens": 880885867.0, "step": 23089 }, { "epoch": 2.9372853326548785, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.692557334899902, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8861765265464783, "num_tokens": 880928408.0, "step": 23090 }, { "epoch": 2.937412542933469, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.6470365524292, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8744121789932251, "num_tokens": 880960681.0, "step": 23091 }, { "epoch": 2.9375397532120595, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.648061752319336, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8621425628662109, "num_tokens": 881000358.0, "step": 23092 }, { "epoch": 2.93766696349065, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.687227249145508, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8753800392150879, "num_tokens": 881037765.0, "step": 23093 }, { "epoch": 2.9377941737692406, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.675324440002441, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8880027532577515, "num_tokens": 881076091.0, "step": 23094 }, { "epoch": 2.937921384047831, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.695487976074219, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.872942328453064, "num_tokens": 881112784.0, "step": 23095 }, { "epoch": 2.9380485943264216, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68996810913086, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8803796768188477, "num_tokens": 881149938.0, "step": 23096 }, { "epoch": 2.938175804605012, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.691215515136719, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8652774095535278, "num_tokens": 881187541.0, "step": 23097 }, { "epoch": 2.9383030148836027, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.655862808227539, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8766821026802063, "num_tokens": 881221920.0, "step": 23098 }, { "epoch": 2.938430225162193, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.677321434020996, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.858736515045166, "num_tokens": 881264804.0, "step": 23099 }, { "epoch": 2.9385574354407837, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.701175689697266, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8657764196395874, "num_tokens": 881304482.0, "step": 23100 }, { "epoch": 2.9386846457193743, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.721339225769043, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.864177942276001, "num_tokens": 881340822.0, "step": 23101 }, { "epoch": 2.938811855997965, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.72094440460205, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8724601864814758, "num_tokens": 881374352.0, "step": 23102 }, { "epoch": 2.9389390662765553, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.69300365447998, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8685674071311951, "num_tokens": 881413176.0, "step": 23103 }, { "epoch": 2.9390662765551454, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.638589859008789, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8692750930786133, "num_tokens": 881454309.0, "step": 23104 }, { "epoch": 2.9391934868337364, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.759784698486328, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8745970726013184, "num_tokens": 881490702.0, "step": 23105 }, { "epoch": 2.9393206971123265, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.660221099853516, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8800351619720459, "num_tokens": 881530394.0, "step": 23106 }, { "epoch": 2.9394479073909174, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.664153099060059, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8876078724861145, "num_tokens": 881567286.0, "step": 23107 }, { "epoch": 2.9395751176695075, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.74488353729248, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8690201044082642, "num_tokens": 881607017.0, "step": 23108 }, { "epoch": 2.939702327948098, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.726180076599121, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8752073049545288, "num_tokens": 881645793.0, "step": 23109 }, { "epoch": 2.9398295382266886, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.637121200561523, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8770379424095154, "num_tokens": 881683596.0, "step": 23110 }, { "epoch": 2.939956748505279, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.714656829833984, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8585902452468872, "num_tokens": 881717607.0, "step": 23111 }, { "epoch": 2.9400839587838696, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.695670127868652, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8867418766021729, "num_tokens": 881750884.0, "step": 23112 }, { "epoch": 2.94021116906246, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.697250366210938, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8693886995315552, "num_tokens": 881789345.0, "step": 23113 }, { "epoch": 2.9403383793410507, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.705554008483887, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8700184226036072, "num_tokens": 881828459.0, "step": 23114 }, { "epoch": 2.940465589619641, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.739449501037598, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8687465190887451, "num_tokens": 881865294.0, "step": 23115 }, { "epoch": 2.9405927998982317, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.6934175491333, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.875782310962677, "num_tokens": 881904936.0, "step": 23116 }, { "epoch": 2.9407200101768223, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.655925750732422, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8852108716964722, "num_tokens": 881936300.0, "step": 23117 }, { "epoch": 2.940847220455413, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.762197494506836, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8852385878562927, "num_tokens": 881974092.0, "step": 23118 }, { "epoch": 2.9409744307340033, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.625852584838867, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8749163746833801, "num_tokens": 882019339.0, "step": 23119 }, { "epoch": 2.941101641012594, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.801356315612793, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.876926064491272, "num_tokens": 882052837.0, "step": 23120 }, { "epoch": 2.9412288512911844, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.726678848266602, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8792209625244141, "num_tokens": 882088449.0, "step": 23121 }, { "epoch": 2.941356061569775, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.794525146484375, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8654050230979919, "num_tokens": 882124844.0, "step": 23122 }, { "epoch": 2.9414832718483654, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.709637641906738, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8769384026527405, "num_tokens": 882162246.0, "step": 23123 }, { "epoch": 2.941610482126956, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.788521766662598, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8628351092338562, "num_tokens": 882202825.0, "step": 23124 }, { "epoch": 2.9417376924055465, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.754591941833496, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.860456645488739, "num_tokens": 882237951.0, "step": 23125 }, { "epoch": 2.941864902684137, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.707627296447754, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8731377124786377, "num_tokens": 882279147.0, "step": 23126 }, { "epoch": 2.941992112962727, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.706656455993652, "learning_rate": 1e-06, "loss": 0.3367, "mean_token_accuracy": 0.8941437005996704, "num_tokens": 882322777.0, "step": 23127 }, { "epoch": 2.942119323241318, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71978759765625, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8796154260635376, "num_tokens": 882360681.0, "step": 23128 }, { "epoch": 2.942246533519908, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.708558082580566, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8548411130905151, "num_tokens": 882403823.0, "step": 23129 }, { "epoch": 2.942373743798499, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.680438041687012, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8821141719818115, "num_tokens": 882443723.0, "step": 23130 }, { "epoch": 2.9425009540770892, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.793014526367188, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8600899577140808, "num_tokens": 882480610.0, "step": 23131 }, { "epoch": 2.94262816435568, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.619431495666504, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8762916326522827, "num_tokens": 882518789.0, "step": 23132 }, { "epoch": 2.9427553746342703, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.768568992614746, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8617397546768188, "num_tokens": 882556598.0, "step": 23133 }, { "epoch": 2.942882584912861, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.73041820526123, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8705841898918152, "num_tokens": 882593788.0, "step": 23134 }, { "epoch": 2.9430097951914513, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.726461410522461, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8805999159812927, "num_tokens": 882630861.0, "step": 23135 }, { "epoch": 2.943137005470042, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.76871395111084, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8823622465133667, "num_tokens": 882670950.0, "step": 23136 }, { "epoch": 2.9432642157486324, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.822647094726562, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8649848103523254, "num_tokens": 882705680.0, "step": 23137 }, { "epoch": 2.943391426027223, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.698599815368652, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8836724162101746, "num_tokens": 882737589.0, "step": 23138 }, { "epoch": 2.9435186363058135, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.811982154846191, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8596585988998413, "num_tokens": 882774502.0, "step": 23139 }, { "epoch": 2.943645846584404, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.65501880645752, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8647714853286743, "num_tokens": 882815184.0, "step": 23140 }, { "epoch": 2.9437730568629945, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.688746452331543, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8862947225570679, "num_tokens": 882856793.0, "step": 23141 }, { "epoch": 2.943900267141585, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.801257133483887, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8845328092575073, "num_tokens": 882896128.0, "step": 23142 }, { "epoch": 2.9440274774201756, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71131706237793, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.881781280040741, "num_tokens": 882931253.0, "step": 23143 }, { "epoch": 2.944154687698766, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.718969345092773, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8903199434280396, "num_tokens": 882975072.0, "step": 23144 }, { "epoch": 2.9442818979773566, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.693816184997559, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8863478302955627, "num_tokens": 883013576.0, "step": 23145 }, { "epoch": 2.944409108255947, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.803223609924316, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8764786720275879, "num_tokens": 883046052.0, "step": 23146 }, { "epoch": 2.9445363185345377, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.6903076171875, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8550573587417603, "num_tokens": 883080398.0, "step": 23147 }, { "epoch": 2.944663528813128, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.771968841552734, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.865428626537323, "num_tokens": 883120321.0, "step": 23148 }, { "epoch": 2.9447907390917187, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.671490669250488, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8807743191719055, "num_tokens": 883161285.0, "step": 23149 }, { "epoch": 2.9449179493703093, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.738997459411621, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.87257981300354, "num_tokens": 883199857.0, "step": 23150 }, { "epoch": 2.9450451596489, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.717938423156738, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8805893659591675, "num_tokens": 883238325.0, "step": 23151 }, { "epoch": 2.94517236992749, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.727568626403809, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8656162023544312, "num_tokens": 883281814.0, "step": 23152 }, { "epoch": 2.945299580206081, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.779540061950684, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8794039487838745, "num_tokens": 883317385.0, "step": 23153 }, { "epoch": 2.945426790484671, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.729720115661621, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8816605806350708, "num_tokens": 883353498.0, "step": 23154 }, { "epoch": 2.945554000763262, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71928596496582, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8828567266464233, "num_tokens": 883391709.0, "step": 23155 }, { "epoch": 2.945681211041852, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.748022079467773, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8588483929634094, "num_tokens": 883426329.0, "step": 23156 }, { "epoch": 2.945808421320443, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.716048240661621, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.872098445892334, "num_tokens": 883465797.0, "step": 23157 }, { "epoch": 2.945935631599033, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.688189506530762, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8763134479522705, "num_tokens": 883501632.0, "step": 23158 }, { "epoch": 2.9460628418776236, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.709211349487305, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8840636014938354, "num_tokens": 883537622.0, "step": 23159 }, { "epoch": 2.946190052156214, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71567440032959, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8761072754859924, "num_tokens": 883576458.0, "step": 23160 }, { "epoch": 2.9463172624348046, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.685158729553223, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8650990724563599, "num_tokens": 883620929.0, "step": 23161 }, { "epoch": 2.946444472713395, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.715010643005371, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8505061268806458, "num_tokens": 883652589.0, "step": 23162 }, { "epoch": 2.9465716829919857, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719019889831543, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.874816358089447, "num_tokens": 883691844.0, "step": 23163 }, { "epoch": 2.946698893270576, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.711617469787598, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8551827669143677, "num_tokens": 883732838.0, "step": 23164 }, { "epoch": 2.9468261035491667, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.711353302001953, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8722639083862305, "num_tokens": 883771415.0, "step": 23165 }, { "epoch": 2.9469533138277573, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.800884246826172, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.859535276889801, "num_tokens": 883808807.0, "step": 23166 }, { "epoch": 2.947080524106348, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.629302978515625, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8882747292518616, "num_tokens": 883844300.0, "step": 23167 }, { "epoch": 2.9472077343849383, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.759199142456055, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8882681131362915, "num_tokens": 883879337.0, "step": 23168 }, { "epoch": 2.947334944663529, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.668489456176758, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8638148903846741, "num_tokens": 883925810.0, "step": 23169 }, { "epoch": 2.9474621549421194, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.605704307556152, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8835970759391785, "num_tokens": 883964753.0, "step": 23170 }, { "epoch": 2.94758936522071, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.768964767456055, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.866950511932373, "num_tokens": 884005737.0, "step": 23171 }, { "epoch": 2.9477165754993004, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.612339973449707, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8749308586120605, "num_tokens": 884043414.0, "step": 23172 }, { "epoch": 2.947843785777891, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.790757179260254, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8785433173179626, "num_tokens": 884077399.0, "step": 23173 }, { "epoch": 2.9479709960564815, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.643173217773438, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.872168779373169, "num_tokens": 884116539.0, "step": 23174 }, { "epoch": 2.948098206335072, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.673540115356445, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8803514838218689, "num_tokens": 884149756.0, "step": 23175 }, { "epoch": 2.9482254166136626, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.728343963623047, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.8934086561203003, "num_tokens": 884181960.0, "step": 23176 }, { "epoch": 2.9483526268922526, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.697887420654297, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8708846569061279, "num_tokens": 884223562.0, "step": 23177 }, { "epoch": 2.9484798371708436, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.703474044799805, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8803088068962097, "num_tokens": 884264541.0, "step": 23178 }, { "epoch": 2.9486070474494337, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.744794845581055, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8615463972091675, "num_tokens": 884301171.0, "step": 23179 }, { "epoch": 2.9487342577280247, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.722450256347656, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8728554248809814, "num_tokens": 884332481.0, "step": 23180 }, { "epoch": 2.9488614680066147, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.72871208190918, "learning_rate": 1e-06, "loss": 0.338, "mean_token_accuracy": 0.8952881097793579, "num_tokens": 884371616.0, "step": 23181 }, { "epoch": 2.9489886782852053, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.772138595581055, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8641031384468079, "num_tokens": 884411245.0, "step": 23182 }, { "epoch": 2.949115888563796, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.677559852600098, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8724822998046875, "num_tokens": 884445712.0, "step": 23183 }, { "epoch": 2.9492430988423863, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.76180648803711, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8780133128166199, "num_tokens": 884484935.0, "step": 23184 }, { "epoch": 2.949370309120977, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.762433052062988, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.87497478723526, "num_tokens": 884518714.0, "step": 23185 }, { "epoch": 2.9494975193995674, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.730669021606445, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8661800622940063, "num_tokens": 884560430.0, "step": 23186 }, { "epoch": 2.949624729678158, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.781448364257812, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8661775588989258, "num_tokens": 884595937.0, "step": 23187 }, { "epoch": 2.9497519399567484, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.728095054626465, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8688679933547974, "num_tokens": 884633201.0, "step": 23188 }, { "epoch": 2.949879150235339, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.725446701049805, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8754619359970093, "num_tokens": 884672602.0, "step": 23189 }, { "epoch": 2.9500063605139295, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.70505142211914, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8783161044120789, "num_tokens": 884708378.0, "step": 23190 }, { "epoch": 2.95013357079252, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.683881759643555, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8718866109848022, "num_tokens": 884747527.0, "step": 23191 }, { "epoch": 2.9502607810711106, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.660951614379883, "learning_rate": 1e-06, "loss": 0.3414, "mean_token_accuracy": 0.8915843367576599, "num_tokens": 884781738.0, "step": 23192 }, { "epoch": 2.950387991349701, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.743159294128418, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8634363412857056, "num_tokens": 884818884.0, "step": 23193 }, { "epoch": 2.9505152016282916, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.700261116027832, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8849878311157227, "num_tokens": 884851751.0, "step": 23194 }, { "epoch": 2.950642411906882, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.696102142333984, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8784686326980591, "num_tokens": 884892935.0, "step": 23195 }, { "epoch": 2.9507696221854727, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.619010925292969, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.87783282995224, "num_tokens": 884935820.0, "step": 23196 }, { "epoch": 2.950896832464063, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.662089347839355, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8728422522544861, "num_tokens": 884976042.0, "step": 23197 }, { "epoch": 2.9510240427426537, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.711261749267578, "learning_rate": 1e-06, "loss": 0.3376, "mean_token_accuracy": 0.8928560614585876, "num_tokens": 885013511.0, "step": 23198 }, { "epoch": 2.9511512530212443, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.67747688293457, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8619078993797302, "num_tokens": 885047291.0, "step": 23199 }, { "epoch": 2.951278463299835, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.728669166564941, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8707736730575562, "num_tokens": 885083116.0, "step": 23200 }, { "epoch": 2.9514056735784253, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.679862976074219, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8702510595321655, "num_tokens": 885123282.0, "step": 23201 }, { "epoch": 2.9515328838570154, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.668333053588867, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8784717321395874, "num_tokens": 885166277.0, "step": 23202 }, { "epoch": 2.9516600941356064, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.668779373168945, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8853321671485901, "num_tokens": 885206414.0, "step": 23203 }, { "epoch": 2.9517873044141965, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.698410034179688, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8824074268341064, "num_tokens": 885240977.0, "step": 23204 }, { "epoch": 2.9519145146927874, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.685249328613281, "learning_rate": 1e-06, "loss": 0.3492, "mean_token_accuracy": 0.8895049095153809, "num_tokens": 885283190.0, "step": 23205 }, { "epoch": 2.9520417249713775, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.702545166015625, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8887242674827576, "num_tokens": 885321340.0, "step": 23206 }, { "epoch": 2.952168935249968, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.732519149780273, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8831795454025269, "num_tokens": 885357524.0, "step": 23207 }, { "epoch": 2.9522961455285586, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.773258209228516, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.851357102394104, "num_tokens": 885389696.0, "step": 23208 }, { "epoch": 2.952423355807149, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.636429786682129, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8890339732170105, "num_tokens": 885429574.0, "step": 23209 }, { "epoch": 2.9525505660857396, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.71412181854248, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8784916996955872, "num_tokens": 885459961.0, "step": 23210 }, { "epoch": 2.95267777636433, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.6259765625, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8679123520851135, "num_tokens": 885506725.0, "step": 23211 }, { "epoch": 2.9528049866429207, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.670913696289062, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8792474269866943, "num_tokens": 885549409.0, "step": 23212 }, { "epoch": 2.952932196921511, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.678336143493652, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8806717395782471, "num_tokens": 885589272.0, "step": 23213 }, { "epoch": 2.9530594072001017, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.660956382751465, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8560205101966858, "num_tokens": 885625951.0, "step": 23214 }, { "epoch": 2.9531866174786923, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.643729209899902, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8772369027137756, "num_tokens": 885665844.0, "step": 23215 }, { "epoch": 2.953313827757283, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.724303245544434, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8734626770019531, "num_tokens": 885705616.0, "step": 23216 }, { "epoch": 2.9534410380358733, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.682254791259766, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8865571618080139, "num_tokens": 885752362.0, "step": 23217 }, { "epoch": 2.953568248314464, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.732666015625, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8700754642486572, "num_tokens": 885792485.0, "step": 23218 }, { "epoch": 2.9536954585930544, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.756077766418457, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8760035634040833, "num_tokens": 885834809.0, "step": 23219 }, { "epoch": 2.953822668871645, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.751129150390625, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.879313588142395, "num_tokens": 885870555.0, "step": 23220 }, { "epoch": 2.9539498791502354, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.702356338500977, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.876861572265625, "num_tokens": 885911911.0, "step": 23221 }, { "epoch": 2.954077089428826, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.720108032226562, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8866528272628784, "num_tokens": 885941094.0, "step": 23222 }, { "epoch": 2.9542042997074165, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.741451263427734, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8736366033554077, "num_tokens": 885978094.0, "step": 23223 }, { "epoch": 2.954331509986007, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.718055725097656, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8909759521484375, "num_tokens": 886014859.0, "step": 23224 }, { "epoch": 2.954458720264597, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.701800346374512, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8808978796005249, "num_tokens": 886048317.0, "step": 23225 }, { "epoch": 2.954585930543188, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.767668724060059, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8778142333030701, "num_tokens": 886081797.0, "step": 23226 }, { "epoch": 2.954713140821778, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.691278457641602, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8617629408836365, "num_tokens": 886124111.0, "step": 23227 }, { "epoch": 2.954840351100369, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.753528594970703, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.885485053062439, "num_tokens": 886162823.0, "step": 23228 }, { "epoch": 2.954967561378959, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.666386604309082, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.878030002117157, "num_tokens": 886204706.0, "step": 23229 }, { "epoch": 2.95509477165755, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.715484619140625, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8837289810180664, "num_tokens": 886241461.0, "step": 23230 }, { "epoch": 2.9552219819361403, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.761129379272461, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8709151744842529, "num_tokens": 886279276.0, "step": 23231 }, { "epoch": 2.955349192214731, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.639373779296875, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8767737150192261, "num_tokens": 886317703.0, "step": 23232 }, { "epoch": 2.9554764024933213, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.706269264221191, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8810437917709351, "num_tokens": 886355235.0, "step": 23233 }, { "epoch": 2.955603612771912, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.694296836853027, "learning_rate": 1e-06, "loss": 0.3255, "mean_token_accuracy": 0.8985270857810974, "num_tokens": 886387354.0, "step": 23234 }, { "epoch": 2.9557308230505024, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.688445091247559, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8723071217536926, "num_tokens": 886419742.0, "step": 23235 }, { "epoch": 2.955858033329093, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.717504501342773, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8766504526138306, "num_tokens": 886455218.0, "step": 23236 }, { "epoch": 2.9559852436076834, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.649282455444336, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8766109347343445, "num_tokens": 886494454.0, "step": 23237 }, { "epoch": 2.956112453886274, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.703256607055664, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8533713817596436, "num_tokens": 886532854.0, "step": 23238 }, { "epoch": 2.9562396641648645, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.768620491027832, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.873344361782074, "num_tokens": 886564288.0, "step": 23239 }, { "epoch": 2.956366874443455, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.680813789367676, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8777695298194885, "num_tokens": 886597155.0, "step": 23240 }, { "epoch": 2.9564940847220456, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.736952781677246, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8782273530960083, "num_tokens": 886632015.0, "step": 23241 }, { "epoch": 2.956621295000636, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.736749649047852, "learning_rate": 1e-06, "loss": 0.3398, "mean_token_accuracy": 0.8926461935043335, "num_tokens": 886665736.0, "step": 23242 }, { "epoch": 2.9567485052792266, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.637726783752441, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8751113414764404, "num_tokens": 886702652.0, "step": 23243 }, { "epoch": 2.956875715557817, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.716081619262695, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8731857538223267, "num_tokens": 886743573.0, "step": 23244 }, { "epoch": 2.9570029258364077, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.707318305969238, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8810356855392456, "num_tokens": 886777932.0, "step": 23245 }, { "epoch": 2.957130136114998, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.75110149383545, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8635572195053101, "num_tokens": 886813211.0, "step": 23246 }, { "epoch": 2.9572573463935887, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.7891263961792, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8796947598457336, "num_tokens": 886849416.0, "step": 23247 }, { "epoch": 2.9573845566721793, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.676323890686035, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8606145977973938, "num_tokens": 886887983.0, "step": 23248 }, { "epoch": 2.95751176695077, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.70547866821289, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8594010472297668, "num_tokens": 886930867.0, "step": 23249 }, { "epoch": 2.95763897722936, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.694405555725098, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8734566569328308, "num_tokens": 886971862.0, "step": 23250 }, { "epoch": 2.957766187507951, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68358039855957, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8818621635437012, "num_tokens": 887005238.0, "step": 23251 }, { "epoch": 2.957893397786541, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.800799369812012, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8688129186630249, "num_tokens": 887035588.0, "step": 23252 }, { "epoch": 2.958020608065132, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.73616886138916, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8763416409492493, "num_tokens": 887079962.0, "step": 23253 }, { "epoch": 2.958147818343722, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.729125022888184, "learning_rate": 1e-06, "loss": 0.3338, "mean_token_accuracy": 0.8943166136741638, "num_tokens": 887114572.0, "step": 23254 }, { "epoch": 2.958275028622313, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.751501083374023, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8791677355766296, "num_tokens": 887149396.0, "step": 23255 }, { "epoch": 2.958402238900903, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.74263858795166, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8697229623794556, "num_tokens": 887191916.0, "step": 23256 }, { "epoch": 2.9585294491794936, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.830941200256348, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8847447633743286, "num_tokens": 887228263.0, "step": 23257 }, { "epoch": 2.958656659458084, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.766722679138184, "learning_rate": 1e-06, "loss": 0.3557, "mean_token_accuracy": 0.8901963829994202, "num_tokens": 887258688.0, "step": 23258 }, { "epoch": 2.9587838697366746, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.754171371459961, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8713662624359131, "num_tokens": 887301565.0, "step": 23259 }, { "epoch": 2.958911080015265, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.787609100341797, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8714091181755066, "num_tokens": 887337511.0, "step": 23260 }, { "epoch": 2.9590382902938557, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.838943481445312, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8792513608932495, "num_tokens": 887370694.0, "step": 23261 }, { "epoch": 2.959165500572446, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.676290512084961, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8740769624710083, "num_tokens": 887412131.0, "step": 23262 }, { "epoch": 2.9592927108510367, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.714447975158691, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8706634044647217, "num_tokens": 887447179.0, "step": 23263 }, { "epoch": 2.9594199211296273, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.753207206726074, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8708375096321106, "num_tokens": 887486079.0, "step": 23264 }, { "epoch": 2.959547131408218, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.708701133728027, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.884152889251709, "num_tokens": 887519092.0, "step": 23265 }, { "epoch": 2.9596743416868083, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.743608474731445, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8780637979507446, "num_tokens": 887552008.0, "step": 23266 }, { "epoch": 2.959801551965399, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.74331283569336, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8715789318084717, "num_tokens": 887585827.0, "step": 23267 }, { "epoch": 2.9599287622439894, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719622611999512, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8733489513397217, "num_tokens": 887628041.0, "step": 23268 }, { "epoch": 2.96005597252258, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.703813552856445, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8789008855819702, "num_tokens": 887667210.0, "step": 23269 }, { "epoch": 2.9601831828011704, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.778040885925293, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8809944987297058, "num_tokens": 887705429.0, "step": 23270 }, { "epoch": 2.960310393079761, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.74172592163086, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8570216298103333, "num_tokens": 887753188.0, "step": 23271 }, { "epoch": 2.9604376033583515, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.767557144165039, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.859910249710083, "num_tokens": 887795907.0, "step": 23272 }, { "epoch": 2.960564813636942, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.723153114318848, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8750109672546387, "num_tokens": 887834859.0, "step": 23273 }, { "epoch": 2.9606920239155325, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.761252403259277, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8711503148078918, "num_tokens": 887879041.0, "step": 23274 }, { "epoch": 2.9608192341941226, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.778788566589355, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8713564872741699, "num_tokens": 887913752.0, "step": 23275 }, { "epoch": 2.9609464444727136, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.751761436462402, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8809367418289185, "num_tokens": 887950946.0, "step": 23276 }, { "epoch": 2.9610736547513037, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.718754768371582, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8730197548866272, "num_tokens": 887993155.0, "step": 23277 }, { "epoch": 2.9612008650298947, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.785871505737305, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.874219536781311, "num_tokens": 888027831.0, "step": 23278 }, { "epoch": 2.9613280753084847, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.752318382263184, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8856042623519897, "num_tokens": 888066941.0, "step": 23279 }, { "epoch": 2.9614552855870753, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.707860946655273, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.890169084072113, "num_tokens": 888099739.0, "step": 23280 }, { "epoch": 2.961582495865666, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.701969146728516, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8765506744384766, "num_tokens": 888139675.0, "step": 23281 }, { "epoch": 2.9617097061442563, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.751005172729492, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.862816333770752, "num_tokens": 888179374.0, "step": 23282 }, { "epoch": 2.961836916422847, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.769274711608887, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.862525463104248, "num_tokens": 888214139.0, "step": 23283 }, { "epoch": 2.9619641267014374, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.692976951599121, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8739274144172668, "num_tokens": 888246365.0, "step": 23284 }, { "epoch": 2.962091336980028, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.761427879333496, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8691425323486328, "num_tokens": 888290903.0, "step": 23285 }, { "epoch": 2.9622185472586184, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.704401969909668, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8722893595695496, "num_tokens": 888331329.0, "step": 23286 }, { "epoch": 2.962345757537209, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.787938117980957, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8683211803436279, "num_tokens": 888372411.0, "step": 23287 }, { "epoch": 2.9624729678157995, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.650774002075195, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8741872310638428, "num_tokens": 888409778.0, "step": 23288 }, { "epoch": 2.96260017809439, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.676420211791992, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8814331293106079, "num_tokens": 888453593.0, "step": 23289 }, { "epoch": 2.9627273883729806, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.731075286865234, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8875817060470581, "num_tokens": 888488245.0, "step": 23290 }, { "epoch": 2.962854598651571, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719781875610352, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.874350905418396, "num_tokens": 888524306.0, "step": 23291 }, { "epoch": 2.9629818089301616, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.69603443145752, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8876690864562988, "num_tokens": 888563791.0, "step": 23292 }, { "epoch": 2.963109019208752, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.730634689331055, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8794196844100952, "num_tokens": 888605021.0, "step": 23293 }, { "epoch": 2.9632362294873427, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.766963005065918, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8759905099868774, "num_tokens": 888647382.0, "step": 23294 }, { "epoch": 2.963363439765933, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.718385696411133, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8746470212936401, "num_tokens": 888682174.0, "step": 23295 }, { "epoch": 2.9634906500445237, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71396541595459, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8774471879005432, "num_tokens": 888724401.0, "step": 23296 }, { "epoch": 2.9636178603231143, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.69030475616455, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8813048601150513, "num_tokens": 888766230.0, "step": 23297 }, { "epoch": 2.963745070601705, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.770160675048828, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.853509783744812, "num_tokens": 888811865.0, "step": 23298 }, { "epoch": 2.9638722808802953, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.670943260192871, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8753393888473511, "num_tokens": 888849549.0, "step": 23299 }, { "epoch": 2.9639994911588854, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.775932312011719, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8771569132804871, "num_tokens": 888881305.0, "step": 23300 }, { "epoch": 2.9641267014374764, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.749960899353027, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8799477219581604, "num_tokens": 888917862.0, "step": 23301 }, { "epoch": 2.9642539117160664, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.711005210876465, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8754551410675049, "num_tokens": 888953883.0, "step": 23302 }, { "epoch": 2.9643811219946574, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719186782836914, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.8925338983535767, "num_tokens": 888991052.0, "step": 23303 }, { "epoch": 2.9645083322732475, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.679795265197754, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8876427412033081, "num_tokens": 889022523.0, "step": 23304 }, { "epoch": 2.964635542551838, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.663203239440918, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8666954040527344, "num_tokens": 889063470.0, "step": 23305 }, { "epoch": 2.9647627528304286, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.681659698486328, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8836686611175537, "num_tokens": 889101160.0, "step": 23306 }, { "epoch": 2.964889963109019, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.683552742004395, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8721583485603333, "num_tokens": 889139128.0, "step": 23307 }, { "epoch": 2.9650171733876096, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.653611183166504, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8530429601669312, "num_tokens": 889175574.0, "step": 23308 }, { "epoch": 2.9651443836662, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.75507640838623, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8732670545578003, "num_tokens": 889216520.0, "step": 23309 }, { "epoch": 2.9652715939447907, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.70427131652832, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.887137770652771, "num_tokens": 889256746.0, "step": 23310 }, { "epoch": 2.965398804223381, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.712453842163086, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.870572566986084, "num_tokens": 889298868.0, "step": 23311 }, { "epoch": 2.9655260145019717, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.740899085998535, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8708254098892212, "num_tokens": 889337758.0, "step": 23312 }, { "epoch": 2.9656532247805623, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68493938446045, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8792431354522705, "num_tokens": 889375308.0, "step": 23313 }, { "epoch": 2.965780435059153, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.752408981323242, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8837040662765503, "num_tokens": 889413735.0, "step": 23314 }, { "epoch": 2.9659076453377433, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.641993522644043, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8780934810638428, "num_tokens": 889451585.0, "step": 23315 }, { "epoch": 2.966034855616334, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.821863174438477, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8588012456893921, "num_tokens": 889487405.0, "step": 23316 }, { "epoch": 2.9661620658949244, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.66374397277832, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8649617433547974, "num_tokens": 889523051.0, "step": 23317 }, { "epoch": 2.966289276173515, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.774609565734863, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.856918454170227, "num_tokens": 889559822.0, "step": 23318 }, { "epoch": 2.9664164864521054, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.711284637451172, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8753882050514221, "num_tokens": 889591598.0, "step": 23319 }, { "epoch": 2.966543696730696, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.692291259765625, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8845810890197754, "num_tokens": 889629787.0, "step": 23320 }, { "epoch": 2.9666709070092865, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.722783088684082, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8817030191421509, "num_tokens": 889668730.0, "step": 23321 }, { "epoch": 2.966798117287877, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.722302436828613, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8740972280502319, "num_tokens": 889705215.0, "step": 23322 }, { "epoch": 2.966925327566467, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.758238792419434, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8764393329620361, "num_tokens": 889738528.0, "step": 23323 }, { "epoch": 2.967052537845058, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.71856689453125, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8786770105361938, "num_tokens": 889771694.0, "step": 23324 }, { "epoch": 2.967179748123648, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.761637687683105, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8881751298904419, "num_tokens": 889806096.0, "step": 23325 }, { "epoch": 2.967306958402239, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.655899047851562, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8740732669830322, "num_tokens": 889848514.0, "step": 23326 }, { "epoch": 2.967434168680829, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.742437362670898, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8952407836914062, "num_tokens": 889889689.0, "step": 23327 }, { "epoch": 2.96756137895942, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.731024742126465, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8785626888275146, "num_tokens": 889925550.0, "step": 23328 }, { "epoch": 2.9676885892380103, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.79312801361084, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8634692430496216, "num_tokens": 889961797.0, "step": 23329 }, { "epoch": 2.967815799516601, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.662484169006348, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.881206750869751, "num_tokens": 889997935.0, "step": 23330 }, { "epoch": 2.9679430097951913, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.731755256652832, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.887172520160675, "num_tokens": 890037900.0, "step": 23331 }, { "epoch": 2.968070220073782, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.740861892700195, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8806616067886353, "num_tokens": 890074814.0, "step": 23332 }, { "epoch": 2.9681974303523724, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.754975318908691, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8879939317703247, "num_tokens": 890107091.0, "step": 23333 }, { "epoch": 2.968324640630963, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.77505874633789, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8846122622489929, "num_tokens": 890140558.0, "step": 23334 }, { "epoch": 2.9684518509095534, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.740325927734375, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8764796257019043, "num_tokens": 890174563.0, "step": 23335 }, { "epoch": 2.968579061188144, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.697729110717773, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8727787137031555, "num_tokens": 890210584.0, "step": 23336 }, { "epoch": 2.9687062714667345, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.692055702209473, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.888060450553894, "num_tokens": 890251955.0, "step": 23337 }, { "epoch": 2.968833481745325, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.75636100769043, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8713123202323914, "num_tokens": 890289095.0, "step": 23338 }, { "epoch": 2.9689606920239155, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.767766952514648, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.878237247467041, "num_tokens": 890327864.0, "step": 23339 }, { "epoch": 2.969087902302506, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.735280990600586, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8755064010620117, "num_tokens": 890364244.0, "step": 23340 }, { "epoch": 2.9692151125810966, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.708763122558594, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8618822693824768, "num_tokens": 890397652.0, "step": 23341 }, { "epoch": 2.969342322859687, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.75723934173584, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8798958659172058, "num_tokens": 890435598.0, "step": 23342 }, { "epoch": 2.9694695331382777, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.678403854370117, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8790911436080933, "num_tokens": 890473445.0, "step": 23343 }, { "epoch": 2.969596743416868, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.690823554992676, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8800175786018372, "num_tokens": 890514263.0, "step": 23344 }, { "epoch": 2.9697239536954587, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.738988876342773, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8809735774993896, "num_tokens": 890554123.0, "step": 23345 }, { "epoch": 2.9698511639740492, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.704662322998047, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8755502700805664, "num_tokens": 890593938.0, "step": 23346 }, { "epoch": 2.9699783742526398, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.795492172241211, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8659842014312744, "num_tokens": 890632260.0, "step": 23347 }, { "epoch": 2.97010558453123, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.74038314819336, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8729668259620667, "num_tokens": 890666852.0, "step": 23348 }, { "epoch": 2.970232794809821, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.85583209991455, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8714287877082825, "num_tokens": 890707959.0, "step": 23349 }, { "epoch": 2.970360005088411, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.701662063598633, "learning_rate": 1e-06, "loss": 0.4735, "mean_token_accuracy": 0.8552521467208862, "num_tokens": 890748729.0, "step": 23350 }, { "epoch": 2.970487215367002, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.758929252624512, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.8878850936889648, "num_tokens": 890782116.0, "step": 23351 }, { "epoch": 2.970614425645592, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.720626831054688, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8626677989959717, "num_tokens": 890827309.0, "step": 23352 }, { "epoch": 2.970741635924183, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.762718200683594, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8653861284255981, "num_tokens": 890863879.0, "step": 23353 }, { "epoch": 2.970868846202773, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.756438255310059, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8782700896263123, "num_tokens": 890901184.0, "step": 23354 }, { "epoch": 2.9709960564813636, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.725469589233398, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8853791356086731, "num_tokens": 890937779.0, "step": 23355 }, { "epoch": 2.971123266759954, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.762212753295898, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.8875887393951416, "num_tokens": 890978055.0, "step": 23356 }, { "epoch": 2.9712504770385446, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.745926856994629, "learning_rate": 1e-06, "loss": 0.5202, "mean_token_accuracy": 0.8457239270210266, "num_tokens": 891010985.0, "step": 23357 }, { "epoch": 2.971377687317135, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.762582778930664, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.880040168762207, "num_tokens": 891045904.0, "step": 23358 }, { "epoch": 2.9715048975957257, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.691632270812988, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8794147968292236, "num_tokens": 891086240.0, "step": 23359 }, { "epoch": 2.971632107874316, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.709774017333984, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8732336759567261, "num_tokens": 891126482.0, "step": 23360 }, { "epoch": 2.9717593181529067, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.734878540039062, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8652727603912354, "num_tokens": 891165982.0, "step": 23361 }, { "epoch": 2.9718865284314973, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.695599555969238, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8622037172317505, "num_tokens": 891199101.0, "step": 23362 }, { "epoch": 2.972013738710088, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.733879089355469, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8747500777244568, "num_tokens": 891231511.0, "step": 23363 }, { "epoch": 2.9721409489886783, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.70257568359375, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8633602857589722, "num_tokens": 891262369.0, "step": 23364 }, { "epoch": 2.972268159267269, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.699942588806152, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8603413701057434, "num_tokens": 891304206.0, "step": 23365 }, { "epoch": 2.9723953695458594, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.670822143554688, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8628813028335571, "num_tokens": 891348006.0, "step": 23366 }, { "epoch": 2.97252257982445, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.76247501373291, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8823872804641724, "num_tokens": 891386851.0, "step": 23367 }, { "epoch": 2.9726497901030404, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.72033977508545, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8853106498718262, "num_tokens": 891423582.0, "step": 23368 }, { "epoch": 2.972777000381631, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.665956497192383, "learning_rate": 1e-06, "loss": 0.3426, "mean_token_accuracy": 0.8896417617797852, "num_tokens": 891463773.0, "step": 23369 }, { "epoch": 2.9729042106602215, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.767388343811035, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8862599730491638, "num_tokens": 891498117.0, "step": 23370 }, { "epoch": 2.973031420938812, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.717142105102539, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8826040029525757, "num_tokens": 891529055.0, "step": 23371 }, { "epoch": 2.9731586312174025, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.696901321411133, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8843090534210205, "num_tokens": 891567370.0, "step": 23372 }, { "epoch": 2.9732858414959926, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719415664672852, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8785648941993713, "num_tokens": 891602737.0, "step": 23373 }, { "epoch": 2.9734130517745836, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.697615623474121, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8677842617034912, "num_tokens": 891639713.0, "step": 23374 }, { "epoch": 2.9735402620531737, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.721012115478516, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.887170672416687, "num_tokens": 891673931.0, "step": 23375 }, { "epoch": 2.9736674723317646, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.772812843322754, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8748083114624023, "num_tokens": 891708015.0, "step": 23376 }, { "epoch": 2.9737946826103547, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.691483497619629, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8718017339706421, "num_tokens": 891746023.0, "step": 23377 }, { "epoch": 2.9739218928889453, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.718899726867676, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8734261989593506, "num_tokens": 891777767.0, "step": 23378 }, { "epoch": 2.974049103167536, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.72754192352295, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8775953054428101, "num_tokens": 891810816.0, "step": 23379 }, { "epoch": 2.9741763134461263, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.736552238464355, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8688294887542725, "num_tokens": 891848810.0, "step": 23380 }, { "epoch": 2.974303523724717, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.676597595214844, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8791897892951965, "num_tokens": 891892553.0, "step": 23381 }, { "epoch": 2.9744307340033074, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.756061553955078, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8856993913650513, "num_tokens": 891927539.0, "step": 23382 }, { "epoch": 2.974557944281898, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.685675621032715, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8793588280677795, "num_tokens": 891959577.0, "step": 23383 }, { "epoch": 2.9746851545604884, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.672513961791992, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8495502471923828, "num_tokens": 892000799.0, "step": 23384 }, { "epoch": 2.974812364839079, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.815874099731445, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8688546419143677, "num_tokens": 892042200.0, "step": 23385 }, { "epoch": 2.9749395751176695, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.680743217468262, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8657801747322083, "num_tokens": 892084838.0, "step": 23386 }, { "epoch": 2.97506678539626, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.653371810913086, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8597674369812012, "num_tokens": 892126442.0, "step": 23387 }, { "epoch": 2.9751939956748505, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.687368392944336, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8650407791137695, "num_tokens": 892165019.0, "step": 23388 }, { "epoch": 2.975321205953441, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.770037651062012, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8774926662445068, "num_tokens": 892202291.0, "step": 23389 }, { "epoch": 2.9754484162320316, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.70789623260498, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8670909404754639, "num_tokens": 892235159.0, "step": 23390 }, { "epoch": 2.975575626510622, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.746149063110352, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.862949013710022, "num_tokens": 892271967.0, "step": 23391 }, { "epoch": 2.9757028367892127, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.710365295410156, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8612772226333618, "num_tokens": 892306230.0, "step": 23392 }, { "epoch": 2.975830047067803, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.818196296691895, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8774598836898804, "num_tokens": 892342118.0, "step": 23393 }, { "epoch": 2.9759572573463937, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.725543975830078, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8799501061439514, "num_tokens": 892380764.0, "step": 23394 }, { "epoch": 2.9760844676249842, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.787670135498047, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8718709945678711, "num_tokens": 892421113.0, "step": 23395 }, { "epoch": 2.9762116779035748, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.742354393005371, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8813077807426453, "num_tokens": 892457512.0, "step": 23396 }, { "epoch": 2.9763388881821653, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.72424602508545, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8796665668487549, "num_tokens": 892493599.0, "step": 23397 }, { "epoch": 2.9764660984607554, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.712698936462402, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8670399188995361, "num_tokens": 892528474.0, "step": 23398 }, { "epoch": 2.9765933087393464, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.62664794921875, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8809983730316162, "num_tokens": 892569591.0, "step": 23399 }, { "epoch": 2.9767205190179364, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.73143482208252, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8799443244934082, "num_tokens": 892607185.0, "step": 23400 }, { "epoch": 2.9768477292965274, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.722824096679688, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8761197328567505, "num_tokens": 892641453.0, "step": 23401 }, { "epoch": 2.9769749395751175, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.716531753540039, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8701154589653015, "num_tokens": 892676802.0, "step": 23402 }, { "epoch": 2.977102149853708, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.699102401733398, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8864490985870361, "num_tokens": 892711694.0, "step": 23403 }, { "epoch": 2.9772293601322986, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.738286018371582, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8659138679504395, "num_tokens": 892752662.0, "step": 23404 }, { "epoch": 2.977356570410889, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.68005084991455, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.877374529838562, "num_tokens": 892790170.0, "step": 23405 }, { "epoch": 2.9774837806894796, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.699627876281738, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8572579026222229, "num_tokens": 892831437.0, "step": 23406 }, { "epoch": 2.97761099096807, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.614572525024414, "learning_rate": 1e-06, "loss": 0.3331, "mean_token_accuracy": 0.8957703113555908, "num_tokens": 892869022.0, "step": 23407 }, { "epoch": 2.9777382012466607, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.679961204528809, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8762567639350891, "num_tokens": 892910738.0, "step": 23408 }, { "epoch": 2.977865411525251, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.697460174560547, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8836621046066284, "num_tokens": 892952055.0, "step": 23409 }, { "epoch": 2.9779926218038417, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.734395980834961, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.868087887763977, "num_tokens": 892985843.0, "step": 23410 }, { "epoch": 2.9781198320824323, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.639504432678223, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8654483556747437, "num_tokens": 893031197.0, "step": 23411 }, { "epoch": 2.978247042361023, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.720067024230957, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8887696862220764, "num_tokens": 893068113.0, "step": 23412 }, { "epoch": 2.9783742526396133, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.703126907348633, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8856163024902344, "num_tokens": 893109148.0, "step": 23413 }, { "epoch": 2.978501462918204, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.718856811523438, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8795956969261169, "num_tokens": 893151406.0, "step": 23414 }, { "epoch": 2.9786286731967944, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.878111839294434, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8789241909980774, "num_tokens": 893190209.0, "step": 23415 }, { "epoch": 2.978755883475385, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71426010131836, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8743458986282349, "num_tokens": 893232949.0, "step": 23416 }, { "epoch": 2.9788830937539754, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.774341583251953, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8754328489303589, "num_tokens": 893273931.0, "step": 23417 }, { "epoch": 2.979010304032566, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.658806800842285, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8560765981674194, "num_tokens": 893317465.0, "step": 23418 }, { "epoch": 2.9791375143111565, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.754575729370117, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8600921034812927, "num_tokens": 893354089.0, "step": 23419 }, { "epoch": 2.979264724589747, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.965576171875, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.873267412185669, "num_tokens": 893388051.0, "step": 23420 }, { "epoch": 2.979391934868337, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.697668075561523, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8698031306266785, "num_tokens": 893426191.0, "step": 23421 }, { "epoch": 2.979519145146928, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.729098320007324, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8679999709129333, "num_tokens": 893462134.0, "step": 23422 }, { "epoch": 2.979646355425518, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.746219635009766, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8698886036872864, "num_tokens": 893503159.0, "step": 23423 }, { "epoch": 2.979773565704109, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.711837768554688, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8584674596786499, "num_tokens": 893538379.0, "step": 23424 }, { "epoch": 2.979900775982699, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.737934112548828, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8634049892425537, "num_tokens": 893580431.0, "step": 23425 }, { "epoch": 2.98002798626129, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.789297103881836, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8687578439712524, "num_tokens": 893622412.0, "step": 23426 }, { "epoch": 2.9801551965398803, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.761207580566406, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8787205219268799, "num_tokens": 893658575.0, "step": 23427 }, { "epoch": 2.980282406818471, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.665011405944824, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8816855549812317, "num_tokens": 893702812.0, "step": 23428 }, { "epoch": 2.9804096170970613, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.751132011413574, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8676838874816895, "num_tokens": 893742785.0, "step": 23429 }, { "epoch": 2.980536827375652, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.740155220031738, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8801375031471252, "num_tokens": 893782310.0, "step": 23430 }, { "epoch": 2.9806640376542424, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.7166109085083, "learning_rate": 1e-06, "loss": 0.3395, "mean_token_accuracy": 0.8915853500366211, "num_tokens": 893822794.0, "step": 23431 }, { "epoch": 2.980791247932833, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.707046508789062, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8799285888671875, "num_tokens": 893870462.0, "step": 23432 }, { "epoch": 2.9809184582114234, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.655952453613281, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8747899532318115, "num_tokens": 893910654.0, "step": 23433 }, { "epoch": 2.981045668490014, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.828725814819336, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8716024160385132, "num_tokens": 893948819.0, "step": 23434 }, { "epoch": 2.9811728787686045, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.704423904418945, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8799628019332886, "num_tokens": 893983325.0, "step": 23435 }, { "epoch": 2.981300089047195, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.78545093536377, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8730290532112122, "num_tokens": 894023712.0, "step": 23436 }, { "epoch": 2.9814272993257855, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.668404579162598, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8829361200332642, "num_tokens": 894059719.0, "step": 23437 }, { "epoch": 2.981554509604376, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.768623352050781, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8762195706367493, "num_tokens": 894100150.0, "step": 23438 }, { "epoch": 2.9816817198829666, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.708208084106445, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.8546939492225647, "num_tokens": 894134336.0, "step": 23439 }, { "epoch": 2.981808930161557, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.71015739440918, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8610612750053406, "num_tokens": 894168972.0, "step": 23440 }, { "epoch": 2.9819361404401477, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.756379127502441, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.871090829372406, "num_tokens": 894204641.0, "step": 23441 }, { "epoch": 2.982063350718738, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.664368629455566, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8774783611297607, "num_tokens": 894248403.0, "step": 23442 }, { "epoch": 2.9821905609973287, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.735333442687988, "learning_rate": 1e-06, "loss": 0.3386, "mean_token_accuracy": 0.8929508924484253, "num_tokens": 894284035.0, "step": 23443 }, { "epoch": 2.9823177712759192, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.704743385314941, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8785169124603271, "num_tokens": 894321804.0, "step": 23444 }, { "epoch": 2.9824449815545098, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.797662734985352, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8830201625823975, "num_tokens": 894355648.0, "step": 23445 }, { "epoch": 2.9825721918331, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.729558944702148, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8757756948471069, "num_tokens": 894392947.0, "step": 23446 }, { "epoch": 2.982699402111691, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.809049606323242, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8663995265960693, "num_tokens": 894433437.0, "step": 23447 }, { "epoch": 2.982826612390281, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.694195747375488, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8794652819633484, "num_tokens": 894481428.0, "step": 23448 }, { "epoch": 2.982953822668872, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.661087036132812, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.874696671962738, "num_tokens": 894518055.0, "step": 23449 }, { "epoch": 2.983081032947462, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.67401123046875, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8697534799575806, "num_tokens": 894562693.0, "step": 23450 }, { "epoch": 2.983208243226053, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.71270751953125, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8746470212936401, "num_tokens": 894604973.0, "step": 23451 }, { "epoch": 2.983335453504643, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.683172225952148, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8801283836364746, "num_tokens": 894645672.0, "step": 23452 }, { "epoch": 2.9834626637832335, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.71245288848877, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8724901676177979, "num_tokens": 894686702.0, "step": 23453 }, { "epoch": 2.983589874061824, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.685023307800293, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8829953670501709, "num_tokens": 894722768.0, "step": 23454 }, { "epoch": 2.9837170843404146, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.720743179321289, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8860338926315308, "num_tokens": 894759216.0, "step": 23455 }, { "epoch": 2.983844294619005, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.662300109863281, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8800687789916992, "num_tokens": 894799691.0, "step": 23456 }, { "epoch": 2.9839715048975957, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.647478103637695, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8718580007553101, "num_tokens": 894835869.0, "step": 23457 }, { "epoch": 2.984098715176186, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.680113792419434, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8537386655807495, "num_tokens": 894875706.0, "step": 23458 }, { "epoch": 2.9842259254547767, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.704801559448242, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8802940249443054, "num_tokens": 894920135.0, "step": 23459 }, { "epoch": 2.9843531357333672, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.67915153503418, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.876483678817749, "num_tokens": 894960825.0, "step": 23460 }, { "epoch": 2.9844803460119578, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.747156143188477, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8764712810516357, "num_tokens": 895001935.0, "step": 23461 }, { "epoch": 2.9846075562905483, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.708244323730469, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8836441040039062, "num_tokens": 895033559.0, "step": 23462 }, { "epoch": 2.984734766569139, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.710424423217773, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8746901154518127, "num_tokens": 895070958.0, "step": 23463 }, { "epoch": 2.9848619768477294, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.728249549865723, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.873798131942749, "num_tokens": 895111533.0, "step": 23464 }, { "epoch": 2.98498918712632, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.659273147583008, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8707500696182251, "num_tokens": 895149861.0, "step": 23465 }, { "epoch": 2.9851163974049104, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.785802841186523, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.881342351436615, "num_tokens": 895192958.0, "step": 23466 }, { "epoch": 2.985243607683501, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.649778366088867, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8847882151603699, "num_tokens": 895226710.0, "step": 23467 }, { "epoch": 2.9853708179620915, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.765228271484375, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8797460794448853, "num_tokens": 895267695.0, "step": 23468 }, { "epoch": 2.985498028240682, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.660691261291504, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8907048106193542, "num_tokens": 895303101.0, "step": 23469 }, { "epoch": 2.9856252385192725, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.703367233276367, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8811537027359009, "num_tokens": 895340437.0, "step": 23470 }, { "epoch": 2.9857524487978626, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.6729736328125, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8895790576934814, "num_tokens": 895377695.0, "step": 23471 }, { "epoch": 2.9858796590764536, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.738211631774902, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8669546842575073, "num_tokens": 895418599.0, "step": 23472 }, { "epoch": 2.9860068693550437, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.739721298217773, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8743709325790405, "num_tokens": 895454832.0, "step": 23473 }, { "epoch": 2.9861340796336346, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.716606140136719, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8581221699714661, "num_tokens": 895490607.0, "step": 23474 }, { "epoch": 2.9862612899122247, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.731619834899902, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8687151670455933, "num_tokens": 895531340.0, "step": 23475 }, { "epoch": 2.9863885001908153, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.680032730102539, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8752142190933228, "num_tokens": 895568688.0, "step": 23476 }, { "epoch": 2.986515710469406, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.763470649719238, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8660017251968384, "num_tokens": 895607316.0, "step": 23477 }, { "epoch": 2.9866429207479963, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.676471710205078, "learning_rate": 1e-06, "loss": 0.3455, "mean_token_accuracy": 0.891796350479126, "num_tokens": 895648694.0, "step": 23478 }, { "epoch": 2.986770131026587, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.707964897155762, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8706662654876709, "num_tokens": 895687996.0, "step": 23479 }, { "epoch": 2.9868973413051774, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.729418754577637, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8794803619384766, "num_tokens": 895726727.0, "step": 23480 }, { "epoch": 2.987024551583768, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.735246658325195, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8796169757843018, "num_tokens": 895760688.0, "step": 23481 }, { "epoch": 2.9871517618623584, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.802656173706055, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8545882701873779, "num_tokens": 895797303.0, "step": 23482 }, { "epoch": 2.987278972140949, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.722214698791504, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8806410431861877, "num_tokens": 895834078.0, "step": 23483 }, { "epoch": 2.9874061824195395, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.701408386230469, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8652085065841675, "num_tokens": 895872640.0, "step": 23484 }, { "epoch": 2.98753339269813, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.725687980651855, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8887526988983154, "num_tokens": 895907646.0, "step": 23485 }, { "epoch": 2.9876606029767205, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.767321586608887, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8846195936203003, "num_tokens": 895944183.0, "step": 23486 }, { "epoch": 2.987787813255311, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.719175338745117, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8792203664779663, "num_tokens": 895983987.0, "step": 23487 }, { "epoch": 2.9879150235339016, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.701887130737305, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8637951612472534, "num_tokens": 896020928.0, "step": 23488 }, { "epoch": 2.988042233812492, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.690459251403809, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8821823596954346, "num_tokens": 896057459.0, "step": 23489 }, { "epoch": 2.9881694440910826, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.714034080505371, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8715453147888184, "num_tokens": 896095049.0, "step": 23490 }, { "epoch": 2.988296654369673, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.702940940856934, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8869130611419678, "num_tokens": 896127433.0, "step": 23491 }, { "epoch": 2.9884238646482637, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.662103652954102, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8812593221664429, "num_tokens": 896165662.0, "step": 23492 }, { "epoch": 2.9885510749268542, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.817590713500977, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8737478256225586, "num_tokens": 896207753.0, "step": 23493 }, { "epoch": 2.9886782852054448, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.70787525177002, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8745512962341309, "num_tokens": 896249105.0, "step": 23494 }, { "epoch": 2.9888054954840353, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.757965087890625, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8751996755599976, "num_tokens": 896291728.0, "step": 23495 }, { "epoch": 2.9889327057626254, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.763449668884277, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8652620315551758, "num_tokens": 896331729.0, "step": 23496 }, { "epoch": 2.9890599160412163, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.781514167785645, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8650652170181274, "num_tokens": 896371804.0, "step": 23497 }, { "epoch": 2.9891871263198064, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.846288681030273, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8735045194625854, "num_tokens": 896408398.0, "step": 23498 }, { "epoch": 2.9893143365983974, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.773785591125488, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8903622627258301, "num_tokens": 896440496.0, "step": 23499 }, { "epoch": 2.9894415468769875, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.790891647338867, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8767787218093872, "num_tokens": 896475212.0, "step": 23500 }, { "epoch": 2.989568757155578, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.756418228149414, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8721568584442139, "num_tokens": 896517122.0, "step": 23501 }, { "epoch": 2.9896959674341685, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.782095909118652, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.882024884223938, "num_tokens": 896557259.0, "step": 23502 }, { "epoch": 2.989823177712759, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.671247482299805, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8664069771766663, "num_tokens": 896600237.0, "step": 23503 }, { "epoch": 2.9899503879913496, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.719128608703613, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8863654732704163, "num_tokens": 896640564.0, "step": 23504 }, { "epoch": 2.99007759826994, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.740582466125488, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8859307169914246, "num_tokens": 896674164.0, "step": 23505 }, { "epoch": 2.9902048085485307, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.733065605163574, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8701052665710449, "num_tokens": 896716613.0, "step": 23506 }, { "epoch": 2.990332018827121, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.79880428314209, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8648765087127686, "num_tokens": 896755586.0, "step": 23507 }, { "epoch": 2.9904592291057117, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.698968887329102, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8740878105163574, "num_tokens": 896797641.0, "step": 23508 }, { "epoch": 2.9905864393843022, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.724440574645996, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8713920712471008, "num_tokens": 896835113.0, "step": 23509 }, { "epoch": 2.9907136496628928, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.811141967773438, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8752457499504089, "num_tokens": 896866496.0, "step": 23510 }, { "epoch": 2.9908408599414833, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.749444007873535, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8726935386657715, "num_tokens": 896907535.0, "step": 23511 }, { "epoch": 2.990968070220074, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.76811695098877, "learning_rate": 1e-06, "loss": 0.3417, "mean_token_accuracy": 0.8912898302078247, "num_tokens": 896948574.0, "step": 23512 }, { "epoch": 2.9910952804986644, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.691707611083984, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8698329329490662, "num_tokens": 896992377.0, "step": 23513 }, { "epoch": 2.991222490777255, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.730997085571289, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8701445460319519, "num_tokens": 897035411.0, "step": 23514 }, { "epoch": 2.9913497010558454, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.739287376403809, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8771408200263977, "num_tokens": 897074661.0, "step": 23515 }, { "epoch": 2.991476911334436, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.759432792663574, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8769038915634155, "num_tokens": 897115550.0, "step": 23516 }, { "epoch": 2.9916041216130265, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.769967079162598, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8688557147979736, "num_tokens": 897155091.0, "step": 23517 }, { "epoch": 2.991731331891617, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.810873031616211, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8700019121170044, "num_tokens": 897196551.0, "step": 23518 }, { "epoch": 2.991858542170207, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.714580535888672, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.885890007019043, "num_tokens": 897231885.0, "step": 23519 }, { "epoch": 2.991985752448798, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.740035057067871, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8752395510673523, "num_tokens": 897267944.0, "step": 23520 }, { "epoch": 2.992112962727388, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.80589771270752, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8776718378067017, "num_tokens": 897307330.0, "step": 23521 }, { "epoch": 2.992240173005979, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.659597396850586, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8780043125152588, "num_tokens": 897344028.0, "step": 23522 }, { "epoch": 2.992367383284569, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.75972843170166, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.867443323135376, "num_tokens": 897391213.0, "step": 23523 }, { "epoch": 2.99249459356316, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.688039779663086, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8688632249832153, "num_tokens": 897428222.0, "step": 23524 }, { "epoch": 2.9926218038417502, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.79547119140625, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8637491464614868, "num_tokens": 897457737.0, "step": 23525 }, { "epoch": 2.992749014120341, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.779958724975586, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.88181072473526, "num_tokens": 897498925.0, "step": 23526 }, { "epoch": 2.9928762243989313, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.74046516418457, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8704245686531067, "num_tokens": 897533157.0, "step": 23527 }, { "epoch": 2.993003434677522, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.67238712310791, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8774999976158142, "num_tokens": 897574159.0, "step": 23528 }, { "epoch": 2.9931306449561124, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.719513893127441, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8674489855766296, "num_tokens": 897606970.0, "step": 23529 }, { "epoch": 2.993257855234703, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.828106880187988, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8528409004211426, "num_tokens": 897647725.0, "step": 23530 }, { "epoch": 2.9933850655132934, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.618144989013672, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8786834478378296, "num_tokens": 897687109.0, "step": 23531 }, { "epoch": 2.993512275791884, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.78040599822998, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8860373497009277, "num_tokens": 897720994.0, "step": 23532 }, { "epoch": 2.9936394860704745, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.77001667022705, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8906487226486206, "num_tokens": 897749855.0, "step": 23533 }, { "epoch": 2.993766696349065, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.701225280761719, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8739098310470581, "num_tokens": 897792846.0, "step": 23534 }, { "epoch": 2.9938939066276555, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.730639457702637, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8776272535324097, "num_tokens": 897836765.0, "step": 23535 }, { "epoch": 2.994021116906246, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.768593788146973, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8752961754798889, "num_tokens": 897880918.0, "step": 23536 }, { "epoch": 2.9941483271848366, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.74233627319336, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8738917708396912, "num_tokens": 897916676.0, "step": 23537 }, { "epoch": 2.994275537463427, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.811306953430176, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8808991312980652, "num_tokens": 897953968.0, "step": 23538 }, { "epoch": 2.9944027477420176, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.752761840820312, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8748945593833923, "num_tokens": 897993055.0, "step": 23539 }, { "epoch": 2.994529958020608, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.848087310791016, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8827254772186279, "num_tokens": 898030202.0, "step": 23540 }, { "epoch": 2.9946571682991987, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.769322395324707, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8705180287361145, "num_tokens": 898067234.0, "step": 23541 }, { "epoch": 2.9947843785777892, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.798952102661133, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8786832094192505, "num_tokens": 898105423.0, "step": 23542 }, { "epoch": 2.9949115888563798, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.737590789794922, "learning_rate": 1e-06, "loss": 0.3329, "mean_token_accuracy": 0.894028902053833, "num_tokens": 898144550.0, "step": 23543 }, { "epoch": 2.99503879913497, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.741745948791504, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8785858154296875, "num_tokens": 898179277.0, "step": 23544 }, { "epoch": 2.995166009413561, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.800054550170898, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8734960556030273, "num_tokens": 898218518.0, "step": 23545 }, { "epoch": 2.995293219692151, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.756490707397461, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8639076948165894, "num_tokens": 898259266.0, "step": 23546 }, { "epoch": 2.995420429970742, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.773199081420898, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8740478754043579, "num_tokens": 898292648.0, "step": 23547 }, { "epoch": 2.995547640249332, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.730463027954102, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8791694641113281, "num_tokens": 898328413.0, "step": 23548 }, { "epoch": 2.995674850527923, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.767685890197754, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8760130405426025, "num_tokens": 898376963.0, "step": 23549 }, { "epoch": 2.995802060806513, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.784520149230957, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8872844576835632, "num_tokens": 898407256.0, "step": 23550 }, { "epoch": 2.9959292710851035, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.759895324707031, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8784967064857483, "num_tokens": 898448800.0, "step": 23551 }, { "epoch": 2.996056481363694, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.756377220153809, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8721131086349487, "num_tokens": 898488912.0, "step": 23552 }, { "epoch": 2.9961836916422846, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.774857521057129, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8689257502555847, "num_tokens": 898530434.0, "step": 23553 }, { "epoch": 2.996310901920875, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.731263160705566, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8765952587127686, "num_tokens": 898565044.0, "step": 23554 }, { "epoch": 2.9964381121994657, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.76148796081543, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8851234912872314, "num_tokens": 898601798.0, "step": 23555 }, { "epoch": 2.996565322478056, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.752973556518555, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8659451603889465, "num_tokens": 898644789.0, "step": 23556 }, { "epoch": 2.9966925327566467, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.655718803405762, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8749094605445862, "num_tokens": 898683928.0, "step": 23557 }, { "epoch": 2.9968197430352372, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.746963500976562, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8709186911582947, "num_tokens": 898726217.0, "step": 23558 }, { "epoch": 2.9969469533138278, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.697677612304688, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.885007381439209, "num_tokens": 898760575.0, "step": 23559 }, { "epoch": 2.9970741635924183, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.817336082458496, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.871964693069458, "num_tokens": 898801232.0, "step": 23560 }, { "epoch": 2.997201373871009, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.629176139831543, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8812710046768188, "num_tokens": 898837174.0, "step": 23561 }, { "epoch": 2.9973285841495994, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.846630096435547, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8662867546081543, "num_tokens": 898875849.0, "step": 23562 }, { "epoch": 2.99745579442819, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.689664840698242, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8839802742004395, "num_tokens": 898909975.0, "step": 23563 }, { "epoch": 2.9975830047067804, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.750277519226074, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8694301247596741, "num_tokens": 898948250.0, "step": 23564 }, { "epoch": 2.997710214985371, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.744954109191895, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8917641043663025, "num_tokens": 898989122.0, "step": 23565 }, { "epoch": 2.9978374252639615, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.838391304016113, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8713446855545044, "num_tokens": 899028715.0, "step": 23566 }, { "epoch": 2.997964635542552, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.802413940429688, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8784452676773071, "num_tokens": 899064101.0, "step": 23567 }, { "epoch": 2.9980918458211425, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.788006782531738, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8794947266578674, "num_tokens": 899101759.0, "step": 23568 }, { "epoch": 2.9982190560997326, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.65843391418457, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8684574961662292, "num_tokens": 899142273.0, "step": 23569 }, { "epoch": 2.9983462663783236, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.851346015930176, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8770614862442017, "num_tokens": 899175878.0, "step": 23570 }, { "epoch": 2.9984734766569137, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.78354549407959, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8862123489379883, "num_tokens": 899208937.0, "step": 23571 }, { "epoch": 2.9986006869355046, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.770026206970215, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8745640516281128, "num_tokens": 899244088.0, "step": 23572 }, { "epoch": 2.9987278972140947, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.818216323852539, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8710721731185913, "num_tokens": 899280111.0, "step": 23573 }, { "epoch": 2.9988551074926852, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.818445205688477, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8752487897872925, "num_tokens": 899310882.0, "step": 23574 }, { "epoch": 2.9989823177712758, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.738911628723145, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8625503778457642, "num_tokens": 899346820.0, "step": 23575 }, { "epoch": 2.9991095280498663, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.80910873413086, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8890813589096069, "num_tokens": 899382044.0, "step": 23576 }, { "epoch": 2.999236738328457, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.78768539428711, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8762868642807007, "num_tokens": 899418389.0, "step": 23577 }, { "epoch": 2.9993639486070474, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.72663402557373, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8737252950668335, "num_tokens": 899458486.0, "step": 23578 }, { "epoch": 2.999491158885638, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.770435333251953, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8742849230766296, "num_tokens": 899498168.0, "step": 23579 }, { "epoch": 2.9996183691642284, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.716351509094238, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8920437097549438, "num_tokens": 899538655.0, "step": 23580 }, { "epoch": 2.999745579442819, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.785472869873047, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8641781210899353, "num_tokens": 899581693.0, "step": 23581 }, { "epoch": 2.9998727897214095, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.733756065368652, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8729717135429382, "num_tokens": 899623817.0, "step": 23582 }, { "epoch": 3.0, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.767645835876465, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8797476291656494, "num_tokens": 899664226.0, "step": 23583 }, { "epoch": 3.0, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.5987625122070312e-05, "step": 23583, "total_flos": 5.62815163329864e+19, "train_loss": 0.4244929082353308, "train_runtime": 47803.2787, "train_samples_per_second": 7.893, "train_steps_per_second": 0.493 } ], "logging_steps": 1, "max_steps": 23583, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 11792, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.62815163329864e+19, "train_batch_size": 4, "trial_name": null, "trial_params": null }